An introduction to Topological Data Analysis: fundamental

and practical aspects for data scientists

Frédéric Chazal and Bertrand Michel

February 26, 2021

Abstract
arXiv:1710.04019v2 [math.ST] 25 Feb 2021

Topological Data Analysis (tda) is a recent and fast growing field providing a set of
new topological and geometric tools to infer relevant features for possibly complex data.
This paper is a brief introduction, through a few selected topics, to basic fundamental and
practical aspects of tda for non experts.

1 Introduction and motivation

Topological Data Analysis (tda) is a recent field that emerged from various works in applied
(algebraic) topology and computational geometry during the first decade of the century. Al-
though one can trace back geometric approaches for data analysis quite far in the past, tda
really started as a field with the pioneering works of Edelsbrunner et al. (2002) and Zomorodian
and Carlsson (2005) in persistent homology and was popularized in a landmark paper in 2009
Carlsson (2009). tda is mainly motivated by the idea that topology and geometry provide a
powerful approach to infer robust qualitative, and sometimes quantitative, information about
the structure of data - see, e.g. Chazal (2017).
tda aims at providing well-founded mathematical, statistical and algorithmic methods to
infer, analyze and exploit the complex topological and geometric structures underlying data that
are often represented as point clouds in Euclidean or more general metric spaces. During the last
few years, a considerable effort has been made to provide robust and efficient data structures and
algorithms for tda that are now implemented and available and easy to use through standard
libraries such as the Gudhi library 1 (C++ and Python) Maria et al. (2014) and its R software
interface Fasy et al. (2014a), Dionysus2 , PHAT3 , DIPHA4 , or Giotto 5 . Although it is still rapidly
evolving, tda now provides a set of mature and efficient tools that can be used in combination
or complementary to other data sciences tools.

The tdapipeline. tda has recently known developments in various directions and applica-
tion fields. There now exist a large variety of methods inspired by topological and geometric
approaches. Providing a complete overview of all these existing approaches is beyond the scope
of this introductory survey. However, many standard ones rely on the following basic pipeline
that will serve as the backbone of this paper:

1. The input is assumed to be a finite set of points coming with a notion of distance - or sim-
ilarity - between them. This distance can be induced by the metric in the ambient space
(e.g. the Euclidean metric when the data are embedded in Rd ) or come as an intrinsic
1
https://gudhi.inria.fr/
2
http://www.mrzv.org/software/dionysus/
3
https://bitbucket.org/phat-code/phat
4
https://github.com/DIPHA/dipha
5
https://giotto-ai.github.io/gtda-docs/0.4.0/library.html

1
 metric defined by a pairwise distance matrix. The definition of the metric on the data is
usually given as an input or guided by the application. It is however important to notice
that the choice of the metric may be critical to reveal interesting topological and geometric
features of the data.

2. A “continuous” shape is built on top of the data in order to highlight the underlying topology
or geometry. This is often a simplicial complex or a nested family of simplicial complexes,
called a filtration, that reflects the structure of the data at different scales. Simplicial
complexes can be seen as higher dimensional generalizations of neighboring graphs that are
classically built on top of data in many standard data analysis or learning algorithms. The
challenge here is to define such structures that are proven to reflect relevant information
about the structure of data and that can be effectively constructed and manipulated in
practice.
3. Topological or geometric information is extracted from the structures built on top of the
data. This may either results in a full reconstruction, typically a triangulation, of the shape
underlying the data from which topological/geometric features can be easily extracted or,
in crude summaries or approximations from which the extraction of relevant information
requires specific methods, such as e.g. persistent homology. Beyond the identification of
interesting topological/geometric information and its visualization and interpretation, the
challenge at this step is to show its relevance, in particular its stability with respect to
perturbations or presence of noise in the input data. For that purpose, understanding the
statistical behavior of the inferred features is also an important question.
4. The extracted topological and geometric information provides new families of features and
descriptors of the data. They can be used to better understand the data - in particular
through visualization- or they can be combined with other kinds of features for further anal-
ysis and machine learning tasks. These information can also be used to design well-suited
data analysis and machine learning models. Showing the added-value and the comple-
mentarity (with respect to other features) of the information provided by tda tools is an
important question at this step.

tda and statistics. Until quite recently, the theoretical aspects of TDA and topological in-
ference mostly relied on deterministic approaches. These deterministic approaches do not take
into account the random nature of data and the intrinsic variability of the topological quantity
they infer. Consequently, most of the corresponding methods remain exploratory, without being
able to efficiently distinguish between information and what is sometimes called the "topological
noise".
A statistical approach to TDA means that we consider data as generated from an unknown
distribution, but also that the inferred topological features by TDA methods are seen as estima-
tors of topological quantities describing an underlying object. Under this approach, the unknown
object usually corresponds to the support of the data distribution (or part of it). The main goals
of a statistical approach to topological data analysis can be summarized as the following list of
problems:
Topic 1: proving consistency and studying the convergence rates of TDA methods.
Topic 2: providing confidence regions for topological features and discussing the significance of
the estimated topological quantities.
Topic 3: selecting relevant scales at which the topological phenomenon should be considered,
as a function of observed data.
Topic 4: dealing with outliers and providing robust methods for TDA.

2
Applications of tda in data science. On the application side, many recent promising and
successful results have demonstrated the interest of topological and geometric approaches in
an increasing number of fields such has, e.g., material science Kramar et al. (2013); Nakamura
et al. (2015) 3D shape analysis Skraba et al. (2010); Turner et al. (2014b), image analysis Qaiser
et al. (2019); Rieck et al. (2020), multivariate time series analysis Khasawneh and Munch (2016);
Seversky et al. (2016); Umeda (2017), medicine Dindin et al. (2020), biology Yao et al. (2009) ,
genomic Carrière and Rabadán (2020) chemistry Lee et al. (2017) or sensor networks De Silva
and Ghrist (2007) to name a few. It is beyond the scope to give an exhaustive list of applications
of tda. On another hand, most of the successes of tda result from its combination with other
analysis or learning techniques - see Section 6.5 for a discussion and references. So, clarifying the
position and complementarity of tda with respect to other approaches and tools in data science
is also an important question and an active research domain.
The overall objective of this survey paper is two-fold. First, it intends to provide data scien-
tists with a brief and comprehensive introduction to the mathematical and statistical foundations
of tda. For that purpose, the focus is put on a few selected, but fundamental, tools and topics:
simplicial complexes (Section 2) and their use for exploratory topological data analysis (Section
3), geometric inference (Section 4) and persistent homology theory (Section 5) that play a central
role in tda. Second, this paper also aims at demonstrating how, thanks to the recent progress of
software, tda tools can be easily applied in data science. In particular, we show how the Python
version of the Gudhi library allows to easily implement and use the tda tools presented in this
paper (Section 7). Our goal is to quickly provide the data scientist with a few basic keys - and
relevant references - to get a clear understanding of the basics of tda to be able to start to use
tda methods and software for his own problems and data.

2 Metric spaces, covers and simplicial complexes

As topological and geometric features are usually associated to continuous spaces, data repre-
sented as finite sets of observations, do not directly reveal any topological information per se.
A natural way to highlight some topological structure out of data is to “connect” data points
that are close to each other in order to exhibit a global continuous shape underlying the data.
Quantifying the notion of closeness between data points is usually done using a distance (or a
dissimilarity measure), and it often turns out to be convenient in tda to consider data sets as
discrete metric spaces or as samples of metric spaces.

Metric spaces. Recall that a metric space (M, ρ) is a set M with a function ρ : M ×M → R+ ,
called a distance, such that for any x, y, z ∈ M :
i) ρ(x, y) ≥ 0 and ρ(x, y) = 0 if and only if x = y,
ii) ρ(x, y) = ρ(y, x) and,
iii) ρ(x, z) ≤ ρ(x, y) + ρ(y, z).
Given a a metric space (M, ρ), the set K(M ) of its compact subsets can be endowed with the so-
called Hausdorff distance: given two compact subsets A, B ⊆ M the Hausdorff distance dH (A, B)
between A and B is defined as the smallest non negative number δ such that for any a ∈ A there
exists b ∈ B such that ρ(a, b) ≤ δ and for any b ∈ B, there exists a ∈ A such that ρ(a, b) ≤ δ - see
Figure 1. In other words, if for any compact subset C ⊆ M , we denote by d(., C) : M → R+ the
distance function to C defined by d(x, C) := inf c∈C ρ(x, c) for any x ∈ M , then one can prove
that the Hausdorff distance between A and B is defined by any of the two following equalities:

dH (A, B) = max{sup d(b, A), sup d(a, B)}

b∈B a∈A
= sup |d(x, A) − d(x, B)| = kd(., A) − d(., B)k∞
x∈M

3
Figure 1: Right: the Hausdorff distance between two subsets A and B of the plane. In this
example, dH (A, B) is the distance between the point a in A which is the farthest from B and its
nearest neighbor b on B. Left: The Gromov-Hausdorff distance between A and B. A can been
rotated - this is an isometric embedding of A in the plane - to reduce its Hausdorff distance to
B. As a consequence, dGH (A, B) ≤ dH (A, B).

It is a basic and classical result that the Hausdorff distance is indeed a distance on the set
of compact subsets of a metric space. From a tda perspective it provides a convenient way to
quantify the proximity between different data sets issued from the same ambient metric space.
However, it sometimes occurs in that one has to compare data set that are not sampled from
the same ambient space. Fortunately, the notion of Hausdorff distance can be generalized to the
comparison of any pair of compact metric spaces, giving rise to the notion of Gromov-Hausdorff
distance.
Two compact metric spaces (M1 , ρ1 ) and (M2 , ρ2 ) are isometric if there exists a bijection
φ : M1 → M2 that preserves distances, i.e. ρ2 (φ(x), φ(y)) = ρ1 (x, y) for any x, y ∈ M1 . The
Gromov-Hausdorff distance measures how far two metric space are from being isometric.
Definition 1. The Gromov-Haudorff distance dGH (M1 , M2 ) between two compact metric spaces
is the infimum of the real numbers r ≥ 0 such that there exists a metric space (M, ρ) and two
compact subspaces C1 , C2 ⊂ M that are isometric to M1 and M2 and such that dH (C1 , C2 ) ≤ r.
The Gromov-Hausdorff distance will be used later, in Section 5, for the study of stability
properties persistence diagrams.
Connecting pairs of nearby data points by edges leads to the standard notion of neighboring
graph from which the connectivity of the data can be analyzed, e.g. using some clustering
algorithms. To go beyond connectivity, a central idea in TDA is to build higher dimensional
equivalent of neighboring graphs by not only connecting pairs but also (k + 1)-uple of nearby
data points. The resulting objects, called simplicial complexes, allow to identify new topological
features such as cycles, voids and their higher dimensional counterpart.

Geometric and abstract simplicial complexes. Simplicial complexes can be seen as higher
dimensional generalization of graphs. They are mathematical objects that are both topological
and combinatorial, a property making them particularly useful for tda.
Given a set X = {x0 , · · · , xk } ⊂ Rd of k + 1 affinely independent points, the k-dimensional
simplex σ = [x0 , · · · xk ] spanned by X is the convex hull of X. The points of X are called the

4
vertices of σ and the simplices spanned by the subsets of X are called the faces of σ. A geometric
simplicial complex K in Rd is a collection of simplices such that:
i) any face of a simplex of K is a simplex of K,
ii) the intersection of any two simplices of K is either empty or a common face of both.
The union of the simplices of K is a subset of Rd called the underlying space of K that inherits
from the topology of Rd . So, K can also be seen as a topological space through its underlying
space. Notice that once its vertices are known, K is fully characterized by the combinatorial
description of a collection of simplices satisfying some incidence rules.
Given a set V , an abstract simplicial complex with vertex set V is a set K̃ of finite subsets
of V such that the elements of V belongs to K̃ and for any σ ∈ K̃ any subset of σ belongs to
K̃. The elements of K̃ are called the faces or the simplices of K̃. The dimension of an abstract
simplex is just its cardinality minus 1 and the dimension of K̃ is the largest dimension of its
simplices. Notice that simplicial complexes of dimension 1 are graphs.
The combinatorial description of any geometric simplicial K obviously gives rise to an abstract
simplicial complex K̃. The converse is also true: one can always associate to an abstract simplicial
complex K̃, a topological space |K̃| such that if K is a geometric complex whose combinatorial
description is the same as K̃, then the underlying space of K is homeomorphic to |K̃|. Such a
K is called a geometric realization of K̃. As a consequence, abstract simplicial complexes can
be seen as topological spaces and geometric complexes can be seen as geometric realizations of
their underlying combinatorial structure. So, one can consider simplicial complexes at the same
time as combinatorial objects that are well-suited for effective computations and as topological
spaces from which topological properties can be inferred.

Building simplicial complexes from data. Given a data set, or more generally a topological
or metric space, there exist many ways to build simplicial complexes. We present here a few
classical examples that are widely used in practice.
A first example, is an immediate extension of the notion of α-neighboring graph. Assume
that we are given a set of points X in a metric space (M, ρ) and a real number α ≥ 0. The
Vietoris-Rips complex Ripsα (X) is the set of simplices [x0 , . . . , xk ] such that dX (xi , xj ) ≤ α for
all (i, j). It follows immediately from the definition that this is an abstract simplicial complex.
However, in general, even when X is a finite subset of Rd , Ripsα (X) does not admit a geometric
realization in Rd ; in particular, it can be of dimension higher than d.
Closely related to the Vietoris-Rips complex is the Čech complex Cechα (X) that is defined
as the set of simplices [x0 , . . . , xk ] such that the k + 1 closed balls B(xi , α) have a non-empty
intersection. Notice that these two complexes are related by

Ripsα (X) ⊆ Cechα (X) ⊆ Rips2α (X)

and that, if X ⊂ Rd then Cechα (X) and Rips2α (X) have the same 1-dimensional skeleton, i.e.
the same set of vertices and edges.

The nerve theorem. The Čech complex is a particular case of a family of complexes associated
to covers. Given a cover U = (Ui )i∈I of M, i.e. a family of sets Ui such that M = ∪i∈I Ui , the
nerve of U is the abstract simplicial complex C(U) whose vertices are the Ui ’s and such that
k
\
σ = [Ui0 , · · · , Uik ] ∈ C(U) if and only if Uij 6= ∅.
j=0

Given a cover of a data set, where each set of the cover can be, for example, a local cluster or
a grouping of data points sharing some common properties, its nerve provides a compact and
global combinatorial description of the relationship between these sets through their intersection
patterns - see Figure 3.

5
Figure 2: The Čech complex Cechα (X) (left) and the and Vietoris-Rips Rips2α (X) (right) of a
finite point cloud in the plane R2 . The bottom part of Cechα (X) is the union of two adjacent
triangles, while the bottom part of Rips2α (X) is the tetrahedron spanned by the four vertices
and all its faces. The dimension of the Čech complex is 2. The dimension of the Vietoris-Rips
complex is 3. Notice that this later is thus not embedded in R2 .

A fundamental theorem in algebraic topology, relates, under some assumptions, the topology
of the nerve of a cover to the topology of the union of the sets of the cover. To be formally
stated, this result, known as the Nerve Theorem, requires to introduce a few notions.
Two topological spaces X and Y are usually considered as being the same from a topological
point of view if they are homeomorphic, i.e. if there exist two continuous bijective maps f : X →
Y and g : Y → X such that f ◦g and g ◦f are the identity map of Y and X respectively. In many
cases, asking X and Y to be homeomorphic turns out to be a too strong requirement to ensure
that X and Y share the same topological features of interest for tda. Two continuous maps
f0 , f1 : X → Y are said to be homotopic is there exists a continuous map H : X × [0, 1] → Y
such that for any x ∈ X, H(x, 0) = f0 (x) and H(x, 1) = g(x). The spaces X and Y are then
said to be homotopy equivalent if there exist two maps f : X → Y and g : Y → X such that
f ◦ g and g ◦ f are homotopic to the identity map of Y and X respectively. The maps f and
g are then called homotopy equivalent. The notion of homotopy equivalence is weaker than the
notion of homeomorphism: if X and Y are homeomorphic then they are obviously homotopy
equivalent, the converse being not true. However, spaces that are homotopy equivalent still share
many topological invariant, in particular they have the same homology - see Section 4.
A space is said to be contractible if it is homotopy equivalent to a point. Basic examples of
contractible spaces are the balls and, more generally, the convex sets in Rd . Open covers whose
all elements and their intersections are contractible have the remarkable following property.

Theorem 1 (Nerve theorem). Let U = (Ui )i∈I be a cover of a topological space X by open sets
such that the intersection of any subcollection of the Ui ’s is either empty or contractible. Then,
X and the nerve C(U) are homotopy equivalent.

It is easy to verify that convex subsets of Euclidean spaces are contractible. As a consequence,
if U = (Ui )i∈I is a collection of convex subsets of Rd then C(U) and ∪i∈I Ui are homotopy
equivalent. In particular, if X is a set of points in Rd , then the Čech complex Cechα (X) is
homotopy equivalent to the union of balls ∪x∈X B(x, α).
The Nerve Theorem plays a fundamental role in tda: it provide a way to encode the topology

6
 Figure 3: The nerve of a cover of a set of sampled points in the plane.

of continuous spaces into abstract combinatorial structures that are well-suited for the design of
effective data structures and algorithms.

3 Using covers and nerves for exploratory data analysis and vi-
sualization: the Mapper algorithm
Using the nerve of covers as a way to summarize, visualize and explore data is a natural idea that
was first proposed for tda in Singh et al. (2007), giving rise to the so-called Mapper algorithm.

Definition 2. Let f : X → Rd , d ≥ 1, be a continuous real valued function and let U = (Ui )i∈I
be a cover of Rd . The pull back cover of X induced by (f, U) is the collection of open sets
(f −1 (Ui ))i∈I . The refined pull back is the collection of connected components of the open sets
f −1 (Ui ), i ∈ I.

The idea of the Mapper algorithm is, given a data set X and well-chosen real valued function
f : X → Rd , to summarize X through the nerve of the refined pull back of a cover U of f (X),
see Figure 4(a) . For well-chosen covers U (see below), this nerve is a graph providing an easy
and convenient way to visualize the summary of the data. It is described in Algorithm 1 and
illustrated on a simple example in Figure 4(b).
The Mapper algorithm is very simple but it raises several questions about the various choices
that are left to the user and that we briefly discuss in the following.

Algorithm 1 The Mapper algorithm

Input: A data set X with a metric or a dissimilarity measure between data points, a function
f : X → R (or Rd ), and a cover U of f (X).
for each U ∈ U, decompose f −1 (U ) into clusters CU,1 , · · · , CU,kU .
Compute the nerve of the cover of X defined by the CU,1 , · · · , CU,kU , U ∈ U
Output: a simplicial complex, the nerve (often a graph for well-chosen covers → easy to
visualize):
- a vertex vU,i for each cluster CU,i ,
- an edge between vU,i and vU 0 ,j iff CU,i ∩ CU 0 ,j 6= ∅

The choice of f . The choice of the function f , sometimes called the filter or lens function,
strongly depends on the features of the data that one expect to highlight. The following ones
are among the ones more or less classically encountered in the literature:

7
Figure 4: (a) The refined pull back cover of the height function on a surface in R3 and its nerve.
(b) The mapper algorithm on a point cloud sampled around a circle.

- Density estimates: the mapper complex may help to understand the structure and connec-
tivity of high density areas (clusters).

- PCA coordinates or coordinates functions obtained from a non linear dimensionality re-
duction (NLDR) technique, eigenfunctions of graph laplacians,... may help to reveal and
understand some ambiguity in the use of non linear dimensionality reductions.

- The centrality function f (x) = y∈X d(x, y) and the eccentricity function f (x) = maxy∈X d(x, y),
P
appears sometimes to be good choices that do not require any specific knowledge about
the data.

- For data that are sampled around 1-dimensional filamentary structures, the distance func-
tion to a given point allows to recover the underlying topology of the filamentary structures
Chazal et al. (2015c).

8
The choice of the cover U. When f is a real valued function, a standard choice is to take
U to be a set of regularly spaced intervals of equal length r > 0 covering the set f (X). The real
r is sometimes called the resolution of the cover and the percentage g of overlap between two
consecutive intervals is called the the gain of the cover. Note that if the gain g is chosen below
50%, then every point of the real line is covered by at most 2 open sets of U and the output nerve
is a graph. It is important to notice that the output of the Mapper is very sensitive to the choice
of U and small changes in the resolution and gain parameters may results in very large changes
in the output, making the method very unstable. A classical strategy consists in exploring some
range of parameters and select the ones that turn out to provide the most informative output
from the user perspective.

The choice of the clusters. The Mapper algorithm requires to cluster the preimage of the
open sets U ∈ U. There are two strategies to compute the clusters. A first strategy consists in
applying, for each U ∈ U, a cluster algorithm, chosen by the user, to the premimage f −1 (U ).
A second, more global, strategy consists in building a neighboring graph on top of the data set
X, e.g. k-NN graph or ε-graph, and, for each U ∈ U, taking the connected components of the
subgraph with vertex set f −1 (U ).

Theoretical and statistical aspects of Mapper. Based on the results on stability and the
structure of Mapper proposed in Carrière and Oudot (2015), advances towards a statistically well-
founded version of Mapper have been obtained recently in Carriere et al. (2018). Unsurprisingly,
the convergence of Mapper depends on both the sampling of the data and the regularity of the
filter function. Moreover, subsampling strategies can be proposed to select a complex in a Rips
filtration at a convenient scale, as well as the resolution and the gain for defining the Mapper
graph. The case of stochastic and multivariate filters has also been studied in Carrière and
Michel (2019). An alternative description of the probabilistic convergence of Mapper, in term
of categorification, has also been proposed in Brown et al. (2020). Other approaches have been
proposed to study and deal with the instabilities of the Mapper algorithm in Dey et al. (2016,
2017).

Data Analysis with Mapper. As an exploratory data analysis tool, Mapper has been suc-
cessfully used for clustering and feature selection. The idea is to identify specific structures in
the Mapper graph (or complex), in particular loops and flares. These structures are then used
to identify interesting clusters or to select features or variable that best discriminate the data
in these structures. Applications on real data, illustrating these techniques, may be found, for
example, in Carrière and Rabadán (2020); Lum et al. (2013); Yao et al. (2009).

4 Geometric reconstruction and homology inference

Another way to build covers and use their nerves to exhibit the topological structure of data is
to consider union of balls centered on the data points. In this section, we assume that Xn =
{x0 , · · · , xn } is a subset of Rd sampled i.i.d. according to a probability measure µ with compact
support M ⊂ Rd . The general strategy to infer topological information about M from µ proceeds
in two steps that are discussed in the following of this section:

1. Xn is covered by a union of balls of fixed radius centered on the xi ’s. Under some regularity
assumptions on M , one can relate the topology of this union of balls to the one of M ;

2. From a practical and algorithmic perspective, topological features of M are inferred from
the nerve of the union of balls, using the Nerve Theorem.

9
 In this framework, it is indeed possible to compare spaces through isotopy equivalence, a
stronger notion than homeomorphism: X ⊆ Rd and Y ⊆ Rd are said to be (ambient) isotopic if
there exists a continuous family of homeomorphisms H : [0, 1] × Rd → Rd , H continuous, such
that for any t ∈ [0, 1], Ht = H(t, .) : Rd → Rd is an homeomorphism, H0 is the identity map
in Rd and H1 (X) = Y . Obviously, if X and Y are isotopic, then they are homeomorphic. The
converse is not true: a knotted and an unknotted circles in R3 are not homeomorphic (notice
that although this claim seems rather intuitive, its formal proof requires the use of some non
obvious algebraic topology tools).

4.1 Distance-like functions and reconstruction

Given a compact subset K of Rd , and a non negative real number r, the union of balls of radius
r centered on K, K r = ∪x∈K B(x, r), called the r-offset of K, is the r-sublevel set of the distance
function dK : Rd → R defined by dK (x) = inf y∈K kx − yk; in other words, K r = d−1 k ([0, r]). This
remark allows to use differential properties of distance functions and to compare the topology of
the offsets of compact sets that are close to each other with respect to the Hausdorff distance.

Definition 3 (Hausdorff distance in Rd ). The Hausdorff distance between two compact subsets
K, K 0 of Rd is defined by

dH (K, K 0 ) = kdK − dK 0 k∞ = sup |dK (x) − dK 0 (x)|.

x∈Rd

In our setting, the considered compact sets are the data set Xn and of the support M of the
measure µ. When M is a smooth compact submanifold, under mild conditions on dH (Xn , M ), for
some well-chosen r, the offsets of Xn are homotopy equivalent to M , Chazal and Lieutier (2008a);
Niyogi et al. (2008) - see Figure 5 for an illustration. These results extend to larger classes of
compact sets and leads to stronger results on the inference of the isotopy type of the offsets of
M , Chazal et al. (2009c,d). They also lead to results on the estimation of other geometric and
differential quantities such as normals Chazal et al. (2009c), curvatures Chazal et al. (2008) or
boundary measures Chazal et al. (2010) under assumptions on the Haussdorff distance between
the underlying shape and the data sample.

Figure 5: The example of a point cloud Xn sampled on the surface of a torus in R3 (top left)
and its offsets for different values of radii r1 < r2 < r3 . For well chosen values of the radius (e.g.
r1 and r2 ), the offsets are clearly homotopy equivalent to a torus.

10
 These results rely on the 1-semiconcavity of the squared distance function d2K , i.e. the
convexity of the function x → kxk2 − d2K (x), and can be naturally stated in the following general
framework.
Definition 4. A function φ : Rd → R+ is distance-like if it is proper (the pre-image of any
compact set in R is a compact set in Rd ) and x → kxk2 − φ2 (x) is convex.
Thanks to its semiconcavity, a distance-like function φ have a well-defined, but not continu-
ous, gradient ∇φ : Rd → Rd that can be integrated into a continuous flow (Petrunin, 2007) that
allows to track the evolution of the topology of its sublevel sets and to compare it to the one of
the sublevel sets of close distance-like functions.
Definition 5. Let φ be a distance-like function and let φr = φ−1 ([0, r]) be the r-sublevel set of
φ.
• A point x ∈ Rd is called α-critical if k∇x φk ≤ α. The corresponding value r = φ(x) is also
said to be α-critical.
• The weak feature size of φ at r is the minimum r0 > 0 such that φ does not have any
critical value between r and r + r0 . We denote it by wfsφ (r). For any 0 < α < 1, the
α-reach of φ is the maximum r such that φ−1 ((0, r]) does not contain any α-critical point.
The weak feature size wfsφ (r) (resp. α-reach) measures the regularity of φ around its r-level
sets (resp. O-level set). When φ = dK is the distance function to a compact set K ⊂ Rd ,
the 1-reach coincides with the classical reach from geometric measure theory Federer (1959).
Its estimation from random samples has been studied in Aamari et al. (2019). An important
property of a distance-like function φ is the topology of their sublevel sets φr can only change
when r crosses a 0-critical value.
Lemma 1 (Isotopy Lemma Grove (1993)). Let φ be a distance-like function and r1 < r2 be two
positive numbers such that φ has no 0-critical point, i.e. points x such that ∇φ(x) = 0, in the
subset φ−1 ([r1 , r2 ]). Then all the sublevel sets φ−1 ([0, r]) are isotopic for r ∈ [r1 , r2 ].
As an immediate consequence of the Isotopy Lemma, all the sublevel sets of φ between r
and r + wfsφ (r) have the same topology. Now the following reconstruction theorem from Chazal
et al. (2011b) provides a connection between the topology of the sublevel sets of close distance-like
functions.
Theorem 2 (Reconstruction Theorem). Let φ, ψ be two distance-like functions such that kφ −
ψk∞ < ε, with reachα (φ) ≥ R for some positive ε and α. Then, for every r ∈ [4ε/α2 , R − 3ε]
and every η ∈ (0, R), the sublevel sets ψ r and φη are homotopy equivalent when
R
ε≤ .
5 + 4/α2
Under similar but slightly more technical conditions the Reconstruction Theorem can be
extended to prove that the sublevel sets are indeed homeomorphic and even isotopic Chazal
et al. (2009c, 2008).
Coming back to our setting, and taking for φ = dM and ψ = dXn the distance functions to
the support M of the measure µ and to the data set Xn , the condition reachα (dM ) ≥ R can
be interpreted as regularity condition on M 6 . The Reconstruction Theorem combined with the
Nerve Theorem tell that, for well-chosen values of r, η, the η-offsets of M are homotopy equivalent
to the nerve of the union of balls of radius r centered on Xn , i.e the Cech complex Cechr (Xn ).
From a statistical perspective, the main advantage of these results involving Hausdorff dis-
tance is that the estimation of the considered topological quantities boil down to support esti-
mation questions that have been widely studied - see Section 4.3.
6
As an example, if M is a smooth compact submanifold then reach0 (φ) is always positive and known as the
reach of M Federer (1959).

11
4.2 Homology inference
The above results provide a mathematically well-founded framework to infer the topology of
shapes from a simplicial complex built on top of an approximating finite sample. However, from
a more practical perspective it raises raise two issues. First, the Reconstruction Theorem requires
a regularity assumption through the α-reach condition that may not always be satisfied and, the
choice of a radius r for the ball used to build the Čech complex Cechr (Xn ). Second, Cechr (Xn )
provides a topologically faithfull summary of the data, through a simplicial complex that is
usually not well-suited for further data processing. One often needs easier to handle topological
descriptors, in particular numerical ones, that can be easily computed from the complex. This
second issue is addressed by considering the homology of the considered simplicial complexes in
the next paragraph, while the first issue will be addressed in the next section with the introduction
of persistent homology.

Homology in a nutshell. Homology is a classical concept in algebraic topology providing a

powerful tool to formalize and handle the notion of topological features of a topological space or
of a simplicial complex in an algebraic way. For any dimension k, the k-dimensional “holes” are
represented by a vector space Hk whose dimension is intuitively the number of such independent
features. For example the 0-dimensional homology group H0 represents the connected compo-
nents of the complex, the 1-dimensional homology group H1 represents the 1-dimensional loops,
the 2-dimensional homology group H2 represents the 2-dimensional cavities,...
To avoid technical subtleties and difficulties, we restrict the introduction of homology to the
minimum that is necessary to understand its usage in the following of the paper. In particular
we restrict to homology with coefficients in Z2 , i.e. the field with two elements 0 and 1 such
that 1 + 1 = 0, that turns out to be geometrically a little bit more intuitive. However, all the
notions and results presented in the sequel naturally extend to homology with coefficient in any
field. We refer the reader to Hatcher (2001) for a complete and comprehensible introduction
to homology and to Ghrist (2017) for a recent concise and very good introduction to applied
algebraic topology and its connections to data analysis.
Let K be a (finite) simplicial complex and let k be a non negative integer. The space of
k-chains on K, Ck (K) is the set whose elements are the formal (finite) sums of k-simplices of K.
More precisely, if {σ1 , · · · σp } is the set of k-simplices of K, then any k-chain can be written as
p
X
c= εi σi with εi ∈ Z2 .
i=1

If c0 = pi=1 ε0i σi is another k-chain and λ ∈ ZP

Pp
2 , the sum c + c is defined as c + c =
0 0
P
i=1 (εi +
p
εi )σi and the product λ.c is defined as λ.c = i=1 (λ.εi )σi , making Ck (K) a vector space with
0

coefficients in Z2 . Since we are considering coefficient in Z2 , geometrically a k-chain can be seen

as a finite collection of k-simplices and the sum of two k-chains as the symmetric difference of
the two corresponding collections7 .
The boundary of a k-simplex σ = [v0 , · · · , vk ] is the (k − 1)-chain
k
X
∂k (σ) = (−1)i [v0 , · · · , v̂i , · · · , vk ]
i=0

where [v0 , · · · , v̂i , · · · , vk ] is the (k − 1)-simplex spanned by all the vertices except vi 8 . As the
k-simplices form a basis of Ck (K), ∂k extends as a linear map from Ck (K) to Ck−1 (K) called
the boundary operator. The kernel Zk (K) = {c ∈ Ck (K) : ∂k = 0} of ∂k is called the space of
k-cycles of K and the image Bk (K) = {c ∈ Ck (K) : ∃c0 ∈ Ck+1 (K), ∂k+1 (c0 ) = c} of ∂k+1 is
7
Recall that the symmetric difference of two sets A and B is the set A∆B = (A \ B) ∪ (B \ A).
8
Notice that as we are considering coefficients in Z2 , here −1 = 1 and thus (−1)i = 1 for any i.

12
called the space of k-boundaries of K. The boundary operators satisfy the fundamental following
property:
∂k−1 ◦ ∂k ≡ 0 for any k ≥ 1.
In other words, any k-boundary is a k-cycle, i.e. Bk (K) ⊆ Zk (K) ⊆ Ck (K). These notions are
illustrated on Figure 6.

Figure 6: Some examples of chains, cycles and boundaries on a 2-dimensional complex K: c1 , c2

and c4 are 1-cycles; c3 si a 1-chain but not a 1-cycle; c4 is the 1-boundary, namely the boundary
of the 2-chain obtained as the sum of the two triangles surrounded by c4 ; The cycles c1 and c2
span the same element in H1 (K) as their difference is the 2-chain represented by the union of
the triangles surrounded by the union of c1 and c2 .

Definition 6 (Simplicial homology group and Betti numbers). The k th (simplicial) homology
group of K is the quotient vector space

Hk (K) = Zk (K)/Bk (K).

The k th Betti number of K is the dimension βk (K) = dim Hk (K) of the vector space Hk (K).

Two cycles c, c0 ∈ Zk (K) are said to be homologous if they differ by a boundary, i.e. is there
exists a (k + 1)-chain d such that c0 = c + ∂k+1 (d). Two such cycles give rise to the same element
of Hk . In other words, the elements of Hk (K) are the equivalence classes of homologous cycles.
Simplicial homology groups and Betti numbers are topological invariants: if K, K 0 are two
simplicial complexes whose geometric realizations are homotopy equivalent, then their homology
groups are isomorphic and their Betti numbers are the same.
Singular homology is another notion of homology that allows to consider larger classes of
topological spaces. It is defined for any topological space X similarly to simplicial homology
except that the notion of simplex is replaced by the notion of singular simplex which is just any
continuous map σ : ∆k → X where ∆k is the standard k-dimensional simplex. The space of k-
chains is the vector space spanned by the k-dimensional singular simplices and the boundary of a
simplex σ is defined as the (alternated) sum of the restriction of σ to the (k −1)-dimensional faces
of ∆k . A remarkable fact about singular homology it that it coincides with simplicial homology

13
Figure 7: The Betti numbers of the circle (top left), the 2-dimensional sphere (top right) and the
2-dimensional torus (bottom). The blue curves on the torus represent two independent cycles
whose homology class is a basis of its 1-dimensional homology group.

whenever X is homeomorphic to the geometric realization of a simplicial complex. This allows

us, in the sequel of this paper, to indifferently talk about simplicial or singular homology for
topological spaces and simplicial complexes.
Observing, that if f : X → Y is a continuous map, then for any singular simplex σ :
∆k → X in X, f ◦ σ : ∆k → Y is a singular simplex in Y , one easily deduces that continuous
maps between topological spaces canonically induce homomorphisms between their homology
groups. In particular, if f is an homeomorphism or an homotopy equivalence, then it induces
an isomorphism between Hk (X) and Hk (Y ) for any non negative integer k. As an example, it
follows from the Nerve Theorem that for any set of points X ⊂ Rd and any r > 0 the r-offset X r
and the Čech complex Cechr (X) have isomorphic homology groups and the same Betti numbers.
As a consequence, the Reconstruction Theorem 2 leads to the following result on the estima-
tion of Betti numbers.

Theorem 3. Let M ⊂ Rd be a compact set such that reachα (dM ) ≥ R > 0 for some α ∈
R
(0, 1) and let X be a finite set of point such that dH (M, X) = ε < 5+4/α 2 . Then, for every
2
r ∈ [4ε/α , R − 3ε] and every η ∈ (0, R), the Betti numbers of Cechr (X) are the same as the ones
of M η .
In particular, if M is a smooth m-dimensional submanifold of Rd , then βk (Cechr (X)) = βk (M )
for any k = 0, · · · , m.

From a practical perspective, this result raises three difficulties: first, the regularity assump-
tion involving the α-reach of M may be too restrictive; second, the computation of the nerve of
an union of balls requires they use of a tricky predicate testing the emptiness of a finite union
of balls; third the estimation of the Betti numbers relies on the scale parameter r whose choice
may be a problem.
To overcome these issues, Chazal and Oudot (2008) establishes the following result that offers
a solution to the two first problems.

Theorem 4. Let M ⊂ Rd be a compact set such that wfs(M ) = wfsdM (0) ≥ R > 0 and let X be
a finite set of point such that dH (M, X) = ε < 19 wfs(M ). Then for any r ∈ [2ε, 14 (wfs(M ) − ε)]

14
and any η ∈ (0, R),
βk (X η ) = rk (Hk (Ripsr (X)) → Hk (Rips4r (X)))
where rk(Hk (Ripsr (X)) → Hk (Rips4r (X))) denotes the rank of the homomorphism induced by the
(continuous) canonical inclusion Ripsr (X) ,→ Rips4r (X).

Although this result leaves the question of the choice of the scale parameter r open, it is
proven in Chazal and Oudot (2008) that a multiscale strategy whose description is beyond the
scope of this paper provides some help to identify the relevant scales at which Theorem 4 can be
applied.

4.3 Statistical aspects of Homology inference

According to the stability results presented in the previous section, a statistical approach to
topological inference is strongly related to the problem of distribution support estimation and
level sets estimation under the Hausdorff metric. A large number of methods and results are
available for estimating the support of a distribution in statistics. For instance, the Devroye and
Wise estimator (Devroye and Wise, 1980) defined on a sample Xn is also a particular offset of
Xn . The convergence rates of both Xn and the Devroye and Wise estimator to the support of
the distribution for the Hausdorff distance is studied in Cuevas and Rodríguez-Casal (2004) in
Rd . More recently, the minimax rates of convergence of manifold estimation for the Hausdorff
metric, which is particularly relevant for topological inference, has been studied in Genovese
et al. (2012). There is also a large literature about level sets estimation in various metrics (see
for instance Cadre, 2006; Polonik, 1995; Tsybakov et al., 1997) and more particularly for the
Hausdorff metric in Chen et al. (2015). All these works about support and level sets estimation
shine light on the statistical analysis of topological inference procedures.
In the paper Niyogi et al. (2008), it is shown that the homotopy type of Riemannian manifolds
with reach larger than a given constant can be recovered with high probability from offsets of
a sample on (or close to) the manifold. This paper was probably the first attempt to consider
the topological inference problem in terms of probability. The result of Niyogi et al. (2008) is
derived from a retract contraction argument and on tight bounds over the packing number of the
manifold in order to control the Hausdorff distance between the manifold and the observed point
cloud. The homology inference in the noisy case, in the sense the distribution of the observation
is concentrated around the manifold, is also studied in Niyogi et al. (2008, 2011). The assumption
that the geometric object is a smooth Riemannian manifold is only used in the paper to control
in probability the Hausdorff distance between the sample and the manifold, and is not actually
necessary for the "topological part" of the result. Regarding the topological results, these are
similar to those of Chazal et al. (2009d); Chazal and Lieutier (2008b) in the particular framework
of Riemannian manifolds. Starting from the result of Niyogi et al. (2008), the minimax rates of
convergence of the homology type have been studied by Balakrishna et al. (2012) under various
models, for Riemannian manifolds with reach larger than a constant. In contrast, a statistical
version of Chazal et al. (2009d) has not yet been proposed.
More recently, following the ideas of Niyogi et al. (2008), Bobrowski et al. (2014) have pro-
posed a robust homology estimator for the level sets of both density and regression functions,
by considering the inclusion map between nested pairs of estimated level sets (in the spirit of
Theorem 4 above) obtained with a plug-in approach from a kernel estimator.

4.4 Going beyond Hausdorff distance : distance to measure

It is well known that distance-based methods in tda may fail completely in the presence of
outliers. Indeed, adding even a single outlier to the point cloud can change the distance function
dramatically, see Figure 8 for an illustration. To answer this drawback, Chazal et al. (2011b) have
introduced an alternative distance function which is robust to noise, the distance-to-a-measure.

15
Figure 8: The effect of outliers on the sublevel sets of distance functions. Adding just a few
outliers to a point cloud may dramatically change its distance function and the topology of its
offsets.

Given a probability distribution P in Rd and a real parameter 0 ≤ u ≤ 1, the notion of

distance to the support of P may be generalized as the function

δP,u : x ∈ Rd 7→ inf{t > 0 ; P (B(x, t)) ≥ u},

where B(x, t) is the closed Euclidean ball of center x and radius t. To avoid issues due to
discontinuities of the map P → δP,u , the distance-to-measure function (DTM) with parameter
m ∈ [0, 1] and power r ≥ 1 is defined by
 Z m 1/r
1
dP,m,r (x) : x ∈ Rd 7→ r
δP,u (x)du . (1)
m 0

A nice property of the DTM proved in Chazal et al. (2011b) is its stability with respect to
1
perturbations of P in the Wasserstein metric. More precisely, the map P → dP,m,r is m− r -
Lipschitz, i.e. if P and P̃ are two probability distributions on Rd , then
1
kdP,m,r − dP̃ ,m,r k∞ ≤ m− r Wr (P, P̃ ) (2)

where Wr is the Wasserstein distance for the Euclidean metric on Rd , with exponent r9 . This
property implies that the DTM associated to close distributions in the Wasserstein metric have
close sublevel sets. Moreover, when r = 2, the function d2P,m,2 is semiconcave ensuring strong
regularity properties on the geometry of its sublevel sets. Using these properties, Chazal et al.
(2011b) show that, under general assumptions, if P̃ is a probability distribution approximating
P , then the sublevel sets of dP̃ ,m,2 provide a topologically correct approximation of the support
of P .
In practice, the measure P is usually only known through a finite set of observations Xn =
{X1 , . . . , Xn } sampled from P , raising the question of the approximation of the DTM. A natural
idea to estimate the DTM from Xn is to plug the empirical measure Pn instead of P in the
definition of the DTM. This "plug-in strategy" corresponds to computing the distance to the
9
See Villani (2003) for a definition of the Wasserstein distance

16
empirical measure (DTEM). For m = nk , the DTEM satisfies

k
1X
drPn ,k/n,r (x) := kx − Xn kr(j) ,
k
j=1

where kx − Xn k(j) denotes the distance between x and its j-th neighbor in {X1 , . . . , Xn }. This
quantity can be easily computed in practice since it only requires the distances between x and
the sample points. The convergence of the DTEM to the DTM has been studied in Chazal et al.
(2014a) and Chazal et al. (2016b).
The introduction of DTM has motivated further works and applications in various directions
such as topological data analysis (Buchet et al., 2015a), GPS traces analysis (Chazal et al.,
2011a), density estimation (Biau et al., 2011), hypothesis testing Brécheteau et al. (2019), clus-
tering (Chazal et al., 2013b) just to name a few. Approximations, generalizations and variants
of the DTM have also been considered in (Brécheteau et al., 2020; Buchet et al., 2015b; Guibas
et al., 2013; Phillips et al., 2014).

5 Persistent homology
Persistent homology is a powerful tool to compute, study and encode efficiently multiscale topo-
logical features of nested families of simplicial complexes and topological spaces. It does not only
provide efficient algorithms to compute the Betti numbers of each complex in the considered fam-
ilies, as required for homology inference in the previous section, but also encodes the evolution
of the homology groups of the nested complexes across the scales. Ideas and preliminary results
underlying persistent homology theory trace back to the 20th century, in particular in Baran-
nikov (1994); Frosini (1992); Robins (1999). It started to know an important development in its
modern form after the seminal works of Edelsbrunner et al. (2002) and Zomorodian and Carlsson
(2005).

5.1 Filtrations
A filtration of a simplicial complex K is a nested family of subcomplexes (Kr )r∈T , where T ⊆ R,
such that for any r, r0 ∈ T , if r ≤ r0 then Kr ⊆ Kr0 , and K = ∪r∈T Kr . The subset T may be
either finite or infinite. More generally, a filtration of a topological space M is a nested family
of subspaces (Mr )r∈T , where T ⊆ R, such that for any r, r0 ∈ T , if r ≤ r0 then Mr ⊆ Mr0 and,
M = ∪r∈T Mr . For example, if f : M → R is a function, then the family Mr = f −1 ((−∞, r]),
r ∈ R defines a filtration called the sublevel set filtration of f .
In practical situations, the parameter r ∈ T can often be interpreted as a scale parameter
and filtrations classically used in TDA often belong to one of the two following families.

Filtrations built on top of data. Given a subset X of a compact metric space (M, ρ), the
families of Rips-Vietoris complexes (Ripsr (X))r∈R and and Čech complexes (Cechr (X))r∈R are
filtrations 10 . Here, the parameter r can be interpreted as a resolution at which one considers the
data set X. For example, if X is a point cloud in Rd , thanks to the Nerve theorem, the filtration
(Cechr (X))r∈R encodes the topology of the whole family of unions of balls Xr = ∪x∈X B(x, r), as
r goes from 0 to +∞. As the notion of filtration is quite flexible, many other filtrations have
been considered in the literature and can be constructed on top of data, such as e.g. the so
called witness complex popularized in tda by De Silva and Carlsson (2004), the weighted Rips
filtrations Buchet et al. (2015b), or the so-called DTM-filtrations Anai et al. (2020) that allow
to handle data corrupted by noise and outliers.
10
we take here the convention that for r < 0, Ripsr (X) = Cechr (X) = ∅

17
Sublevel sets filtrations. Functions defined on the vertices of a simplicial complex give rise
to another important example of filtration: let K be a simplicial complex with vertex set V and
f : V → R. Then f can be extended to all simplices of K by f ([v0 , · · · , vk ]) = max{f (vi ) : i =
1, · · · , k} for any simplex σ = [v0 , · · · , vk ] ∈ K and the family of subcomplexes Kr = {σ ∈ K :
f (σ) ≤ r} defines a filtration call the sublevel set filtration of f . Similarly, one can define the
upperlevel set filtration of f .
In practice, even if the index set is infinite, all the considered filtrations are built on finite sets
and are indeed finite. For example, when X is finite, the Vietoris-Rips complex Ripsr (X) changes
only at a finite number of indices r. This allows to easily handle them from an algorithmic
perspective.

5.2 Starting with a few examples

Given a filtration Filt = (Fr )r∈T of a simplicial complex or a topological space, the homology
of Fr changes as r increases: new connected components can appear, existing component can
merge, loops and cavities can appear or be filled, etc... Persistent homology tracks these changes,
identifies the appearing features and associates a life time to them. The resulting information is
encoded as a set of intervals called a barcode or, equivalently, as a multiset of points in R2 where
the coordinate of each point is the starting and end point of the corresponding interval.
Before giving formal definitions, we introduce and illustrate persistent homology on a few
simple examples.

Example 1. Let f : [0, 1] → R be the function of Figure 9(a) and let (Fr = f −1 ((−∞, r]))r∈R
be the sublevel set filtration of f . All the sublevel sets of f are either empty or a union of intervals,
so the only non trivial topological information they carry is their 0-dimensional homology, i.e.
their number of connected components. For r < a1 , Fr is empty, but at r = a1 a first connected
component appears in Fa1 . Persistent homology thus registers a1 as the birth time of a connected
component and start to keep track of it by creating an interval starting at a1 . Then, Fr remains
connected until r reaches the value a2 where a second connected component appears. Persistent
homology starts to keep track of this new connected component by creating a second interval
starting at a2 . Similarly, when r reaches a3 , a new connected component appears and persistent
homology creates a new interval starting at a3 . When r reaches a4 , the two connected components
created at a1 and a3 merges together to give a single larger component. At this step, persistent
homology follows the rule that this is the most recently appeared component in the filtration
that dies: the interval started at a3 is thus ended at a4 and a first persistence interval encoding
the lifespan of the component born at a3 is created. When r reaches a5 , as in the previous
case, the component born at a2 dies and the persistent interval (a2 , a5 ) is created. The interval
created at a1 remains until the end of the filtration giving rise to the persistent interval (a1 , a6 )
if the filtration is stopped at a6 , or (a1 , +∞) if r goes to +∞ (notice that in this later case, the
filtration remains constant for r > a6 ). The obtained set of intervals encoding the span life of the
different homological features encountered along the filtration is called the persistence barcode of
f . Each interval (a, a0 ) can be represented by the point of coordinates (a, a0 ) in R2 plane. The
resulting set of points is called the persistence diagram of f . Notice that a function may have
several copies of the same interval in its persistence barcode. As a consequence, the persistence
diagram of f is indeed a multi-set where each point has an integer valued multiplicity. Last, for
technical reasons that will become clear in the next section, one adds to the persistence all the
points of the diagonal ∆ = {(b, d) : b = d} with an infinite multiplicity.

Example 2. Let now f : M → R be the function of Figure 9(b) where M is a 2-dimensional

surface homeomorphic to a torus, and let (Fr = f −1 ((−∞, r]))r∈R be the sublevel set filtration
of f . The 0-dimensional persistent homology is computed as in the previous example, giving
rise to the red bars in the barcode. Now, the sublevel sets also carry 1-dimensional homological

18
Figure 9: (a) Example 1: The persistence barcode and the persistence diagram of a function
f : [0, 1] → R. (b) Example 2: the persistence barcode and the persistence diagram of the height
function (projection on the z-axis) defined on a surface in R3 .

19
features. When r goes through the height a1 , the sublevel sets Fr that were homeomorphic to two
discs become homeomorphic to the disjoint union of a disc and an annulus, creating a first cycle
homologous to σ1 on Figure 9(b). A interval (in blue) representing the birth of this new 1-cycle
is thus started at a1 . Similarly, when r goes through the height a2 a second cycle, homologous
to σ2 is created, giving rise to the start of a new persistent interval. These two created cycles
are never filled (indeed they span H1 (M )) and the corresponding intervals remains until the end
of the filtration. When r reaches a3 , a new cycle is created that is filled and thus dies at a4 ,
giving rise to the persistence interval (a3 , a4 ). So, now, the sublevel set filtration of f gives rise
to two barcodes, one for 0-dimensional homology (in red) and one for 1-dimensional homology
(in blue). As previously, these two barcodes can equivalently be represented as diagrams in the
plane.

Example 3. In this last example we consider the filtration given by a union of growing balls
centered on the finite set of points C in Figure 10. Notice that this is the sublevel set filtration
of the distance function to C, and thanks to the Nerve Theorem, this filtration is homotopy
equivalent to the Čech filtration built on top of C. Figure 10 shows several level sets of the
filtration:
a) For the radius r = 0, the union of balls is reduced to the initial finite set of point, each of them
corresponding to a 0-dimensional feature, i.e. a connected component; an interval is created for
the birth for each of these features at r = 0.
b) Some of the balls started to overlap resulting in the death of some connected components that
get merged together; the persistence diagram keeps track of these deaths, putting an end point
to the corresponding intervals as they disappear.
c) New components have merged giving rise to a single connected component and, so, all the
intervals associated to a 0-dimensional feature have been ended, except the one corresponding
to the remaining components; two new 1-dimensional features, have appeared resulting in two
new intervals (in blue) starting at their birth scale.
d) One of the two 1-dimensional cycles has been filled, resulting in its death in the filtration and
the end of the corresponding blue interval.
e) all the 1-dimensional features have died, it only remains the long (and never dying) red
interval. As in the previous examples, the final barcode can also be equivalently represented as
a persistence diagram where every interval (a, b) is represented by the the point of coordinate
(a, b) in R2 . Intuitively the longer is an interval in the barcode or, equivalently the farther
from the diagonal is the corresponding point in the diagram, the more persistent, and thus
relevant, is the corresponding homological feature across the filtration. Notice also that for a
given radius r, the k-th Betti number of the corresponding union of balls is equal of the number of
persistence intervals corresponding to k-dimensional homological features and containing r. So,
the persistence diagram can be seen as a multiscale topological signature encoding the homology
of the union of balls for all radii as well as its evolution across the values of r.

5.3 Persistent modules and persistence diagrams

Persistent diagrams can be formally and rigorously defined in a purely algebraic way. This
requires some care and we only give here the basic necessary notions, leaving aside technical
subtelties and difficulties. We refer the readers interested in a detailed exposition to Chazal
et al. (2016a).
Let Filt = (Fr )r∈T be a filtration of a simplicial complex or a topological space. Given a
non negative integer k and considering the homology groups Hk (Fr ) we obtain a sequence of
vector spaces where the inclusions Fr ⊂ Fr0 , r ≤ r0 induce linear maps between Hk (Fr ) and
Hk (Fr0 ). Such a sequence of vector spaces together with the linear maps connecting them is
called a persistence module.

20
Figure 10: The sublevel set filtration of the distance function to a point cloud and the “construc-
tion” of its persistence barcode as the radius of balls increases. The blue curves in the unions of
balls represent 1-cycles associated to the blue bars in the barcodes.

21
Definition 7. A persistence module V over a subset T of the real numbers R is an indexed family
of vector spaces (Vr | r ∈ T ) and a doubly-indexed family of linear maps (vsr : Vr → Vs | r ≤ s)
which satisfy the composition law vts ◦ vsr = vtr whenever r ≤ s ≤ t, and where vrr is the identity
map on Vr .
In many cases, a persistence module can be decomposed into a direct sum of intervals modules
I(b,d) of the form
· · · → 0 → · · · → 0 → Z2 → · · · → Z2 → 0 → · · ·
where the maps Z2 → Z2 are identity maps while all the other maps are 0. Denoting b (resp.
d) the infimum (resp. supremum) of the interval of indices corresponding to non zero vector
spaces, such a module can be interpreted as a feature that appears in the filtration at index b
and disappear at index d. When a persistence module V can be decomposed as a direct sum of
interval modules, one can show that this decomposition is unique up to reordering the intervals
(see (Chazal et al., 2016a, Theorem 2.7)). As a consequence, the set of resulting intervals is
independent of the decomposition of V and is called the persistence barcode of V. As in the
examples of the previous section, each interval (b, d) in the barcode can be represented as the
point of coordinates (b, d) in the plane R2 . The disjoint union of these points, together with the
diagonale ∆ = {x = y} is a multi-set called the the persistence diagram of V.
The following result, from (Chazal et al., 2016a, Theorem 2.8), gives some necessary condi-
tions for a persistence module to be decomposable as a direct sum of interval modules.
Theorem 5. Let V be a persistence module indexed by T ⊂ R. If T is a finite set or if all
the vector spaces Vr are finite-dimensional, then V is decomposable as a direct sum of interval
modules. Moreover, for any s, t ∈ T , s ≤ t, the number βts of intervals starting before s and
ending after t is equal to the rank of the linear map vts and is called the (s, t)-persistent Betti
number of the filtration.
As both conditions above are satisfied for the persistent homology of filtrations of finite
simplicial complexes, an immediate consequence of this result is that the persistence diagrams
of such filtrations are always well-defined.
Indeed, it is possible to show that persistence diagrams can be defined as soon as the following
simple condition is satisfied.
Definition 8. A persistence module V indexed by T ⊂ R is q-tame if for any r < s in T , the
rank of the linear map vsr : Vr → Vs is finite.
Theorem 6 (Chazal et al. (2009a, 2016a)). If V is a q-tame persistence module, then it has a
well-defined persistence diagram. Such a persistence diagram dgm(V) is the union of the points
of the diagonal ∆ of R2 , counted with infinite multiplicity, and a multi-set above the diagonal
in R2 that is locally finite. Here, by locally finite we mean that for any rectangle R with sides
parallel to the coordinate axes that does not intersect ∆, the number of points of dgm(V), counted
with multiplicity, contained in R is finite. Also, the part of the diagram made of the points with
infinite second coordinate is called the essential part of the diagram.
The construction of persistence diagrams of q-tame modules is beyond the scope of this
paper but it gives rise to the same notion as in the case of decomposable modules. It can be
done either by following the algebraic approach based upon the decomposability properties of
modules, or by adopting a measure theoretic approach that allows to define diagrams as integer
valued measures on a space of rectangles in the plane. We refer the reader to Chazal et al.
(2016a) for more informations.
Although persistence modules encountered in practice are decomposable, the general frame-
work of q-tame persistence module plays a fundamental role in the mathematical and statistical
analysis of persistent homology. In particular, it is needed to ensure the existence of limit dia-
grams when convergence properties are studied - see Section 6.

22
 A filtration Filt = (Fr )r∈T of a simplicial complex or of a topological space is said to be tame
if for any integer k, the persistence module (Hk (Fr ) | r ∈ T ) is q-tame. Notice that the filtrations
of finite simplicial complexes are always tame. As a consequence, for any integer k a persistence
diagram denoted dgmk (Filt) is associated to the filtration Filt. When k is not explicitly specified
and when there is no ambiguity, it is usual to drop the index k in the notation and to talk about
“the” persistence diagram dgm(Filt) of the filtration Filt. This notation has to be understood as
“dgmk (Filt) for some k”.

5.4 Persistence landscapes

The persistence landscape introduced in Bubenik (2015) is an alternative representation of per-
sistence diagrams. This approach aims at representing the topological information encoded in
persistence diagrams as elements of an Hilbert space, for which statistical learning methods can
be directly applied. The persistence landscape is a collection of continuous, piecewise linear
functions λ : N × R → R that summarizes a persistence diagram dgm.

Figure 11: An example of persistence landscape (right) associated to a persistence diagram (left).
The first landscape is in blue, the second one in red and the last one in orange. All the other
landscapes are zero.

A birth-death pair p = (b, d) ∈ dgm is transformed into the point b+d2 , 2 , see Figure 11.
d−b

Remember that the points with infinite persistence have been simply discarded in this definition.
The landscape is then defined by considering the set of functions created by tenting the features
of the rotated persistence diagram as follows:

b+d
t − b t ∈ [b, 2 ]

Λp (t) = d − t t ∈ ( b+d2 , d]
(3)
otherwise.

0


The persistence landscape λdgm of dgm is a summary of the arrangement of piecewise linear
curves obtained by overlaying the graphs of the functions {Λp }p∈dgm . Formally, the persistence
landscape of dgm is the collection of functions

λdgm (k, t) = kmax Λr (t), t ∈ [0, T ], k ∈ N, (4)

r∈dgm

where kmax is the kth largest value in the set; in particular, 1max is the usual maximum function.
Given k ∈ N, the function λdgm (k, .) : R → R is called the k-th landscape of dgm. It is
not difficult to see that the map that associate to each persistence diagram its corresponding
landscape is injective. In other words, formally no information is lost when a persistence diagram
is represented through its persistence landscape.
The advantage of the persistence landscape representation is two-fold. First, persistence di-
agrams are mapped as elements of a functional space, opening the door to the use of a broad

23
variety of statistical and data analysis tools for further processing of topological features - see, e.g.
Bubenik (2015); Chazal et al. (2015b) and Section 6.3.1. Second, and fundamental from a theo-
retical perspective, the persistence landscapes share the same stability properties as persistence
diagrams - see Section 5.7.

5.5 Linear representations of persistence homology

A persistence diagram without its essential part can be represented as a discrete measure on
∆+ = {p = (b, d), b < d < ∞}. With a slight abuse of notation, we can write :
X
dgm = δp ,
p∈dgm

where the features are counted with multiplicity and where δ(b,d) denotes the Dirac measure
in p = (b, d). Most of the persistence-based descriptors that have been proposed to analyze
persistence can be expressed as linear transformations of the persistence diagram, seen as a
point process X
Ψ(dgm) = f (p),
p∈dgm

for some function f defined on ∆ and taking values in a Banach space.

In most cases, we want these transformations to apply independently at each homological
dimension. For k ∈ N a given homological dimension we then consider some linear transformation
of the persistence diagram restricted to the topological features of dimension k:
X
Ψk (dgmk ) = fk (p), (5)
p∈dgmk

where dgmk is the persistence diagram of the topological features of dimension k and where fk
is defined on ∆ and takes values in a Banach space.

Betti curve. The simplest way to represent persistence homology is the Betti function or Betti
curve. The Betti curve of homological dimension k is defined as
X
βk (t) = w(b, d)1t∈[b,d]
(b,d)∈dgm

where w is a weight function defined on ∆. In other words, the Betti curve is the number
of barcodes at time m. This descriptor is a linear representation of persistence homology by
taking f in (5) such that f (b, d)(t) = w(b, d)1t∈[b,d] . A typical choice for the weigh function is
a increasing function of the persistence w(b, d) = w̃(d − b) where w̃ is an increasing function
defined on R+ . One of the first applications of Betti curves can be found in Umeda (2017).

Persistence surface. Persistence surface (also called persistence images) is obtained by mak-
ing the convolution of a diagram with a kernel. It has been introduced in Adams et al. (2017).
For K : R2 → R a kernel and H a 2 × 2 bandwidth matrix (e.g. a symmetric positive definite
matrix), let for u ∈ R2
KH (u) = det(H)−1/2 K(H −1/2 u).
Let w : R2 → R+ a weight function defined on ∆. One defines the persistence surface of
homological dimension k associated to a diagram dgm, with kernel K and bandwidth matrix H
by: X
∀u ∈ R2 , ρk (dgm)(u) = w(r)KH (u − p).
p∈dgmk

The persistence surface is obviously a linear representation of persistence homology. Typical

weigh functions are increasing functions of the persistence.

24
Other linear representations of persistence. Many other linear representations of per-
sistence have been proposed in the literature, such as e. g., the persistence silhouette (Chazal
et al., 2015b), the accumulated persistence function (Biscio and Møller, 2019) and variants of
the persistence surface (Chen et al., 2017; Kusano et al., 2016; Reininghaus et al., 2015).
Considering persistence diagrams as discrete measure and their vectorizations as linear rep-
resentation is an approach that also proven fruitful to study distributions of diagrams Divol and
Chazal (2020) and the metric structure of the space of persistence diagrams Divol and Lacombe
(2020) - see Sections 5.6 and 6.3.

5.6 Metrics on the space of persistence diagrams

To exploit the topological information and topological features inferred from persistent homol-
ogy, one needs to be able to compare persistence diagrams, i.e. to endow the space of persistence
diagrams with a metric structure. Although several metrics can be considered, the most funda-
mental one is known as the bottleneck distance.
Recall that a persistence diagram is the union of a discrete multi-set in the half-plane above
the diagonal ∆ and, for technical reasons that will become clear below, of ∆ where the point
of ∆ are counted with infinite multiplicity. A matching - see Figure 12 - between two diagrams
dgm1 and dgm2 is a subset m ⊆ dgm1 × dgm2 such that every points in dgm1 \∆ and dgm2 \∆
appears exactly once in m. In other words, for any p ∈ dgm1 \∆, and for any q ∈ dgm2 \∆,
({p} × dgm2 ) ∩ m and (dgm1 ×{q}) ∩ m each contains a single pair. The Bottleneck distance
between dgm1 and dgm2 is then defined by

db (dgm1 , dgm2 ) = inf max kp − qk∞ .

matching m (p,q)∈m

Figure 12: A perfect matching and the Bottleneck distance between a blue and a red diagram.
Notice that some points of both diagrams are matched to points of the diagonal.

The practical computation of the bottleneck distance boils down to the computation of a
perfect matching in a bipartite graph for which classical algorithms can be used.
The bottleneck metric is a L∞ -like metric. It turns out to be the natural one to express
stability properties of persistence diagrams presented in Section 5.7, but it suffers from the same
drawbacks as the usual L∞ norms, i.e. it is completely determined by the largest distance
among the pairs and do not take into account the closeness of the remaining pairs of points. A

25
variant, to overcome this issue, the so-called Wasserstein distance between diagrams is sometimes
considered. Given p ≥ 1, it is defined by
X
Wp (dgm1 , dgm2 )p = inf kp − qkp∞ .
matching m
(p,q)∈m

Useful stability results for persistence in the Wp metric exist among the literature, in particular
Cohen-Steiner et al. (2010), but they rely on assumptions that make them consequences of the
stability results in the bottleneck metric. A general study of the space of persistence diagrams
endowed with Wp metrics has been considered in Divol and Lacombe (2020) where they propose
a general framework, based upon optimal partial transport, in which many important properties
of persistence diagrams can be proven in a natural way.

5.7 Stability properties of persistence diagrams

A fundamental property of persistence homology is that persistence diagrams of filtrations built
on top of data sets turn out to be very stable with respect to some perturbations of the data. To
formalize and quantify such stability properties, we first need to precise the notion of perturbation
that are allowed.
Rather than working directly with filtrations built on top of data sets, it turns out to be more
convenient to define a notion of proximity between persistence module from which we will derive
a general stability result for persistent homology. Then, most of the stability results for specific
filtrations will appear as a consequence of this general theorem. To avoid technical discussions,
from now on we assume, without loss of generality, that the considered persistence modules are
indexed by R.

Definition 9. Let V, W be two persistence modules indexed by R. Given δ ∈ R, a homomorphism

of degree δ between V and W is a collection Φ of linear maps φr : Vr → Wr+δ , for all r ∈ R such
r+δ
that for any r ≤ s, φs ◦ vsr = ws+δ ◦ φr .

An important example of homomorphism of degree δ is the shift endomorphism 1δV which

consists of the families of linear maps (vr+δ
r ). Notice also that homomorphisms of modules can

naturally be composed: the composition of a homomorphism Ψ of degree δ between U and V

and a homomorphism Φ of degree δ 0 between V and W naturally gives rise to a homomorphism
ΦΨ of degree δ + δ 0 between U and W.

Definition 10. Let δ ≥ 0. Two persistence modules V, W are δ-interleaved if there exists two
homomorphism of degree δ, Φ, from V to W and Ψ, from W to V such that ΨΦ = 12δ V and
ΦΨ = 12δ
W .

Although it does not define a metric on the space of persistence modules, the notion of
closeness between two persistence module may be defined as the smallest non negative δ such
that they are δ-interleaved. Moreover, it allows to formalize the following fundamental theorem
Chazal et al. (2009a, 2016a).

Theorem 7 (Stability of persistence). Let V and W be two q-tame persistence modules. If V

and W are δ-interleaved for some δ ≥ 0, then

db (dgm(V), dgm(W)) ≤ δ.

Although purely algebraic and rather abstract, this result is a efficient tool to easily establish
concrete stability results in TDA. For example we can easily recover the first persistence stability
result that appeared in the literature (Cohen-Steiner et al., 2005).

26
Theorem 8. Let f, g : M → R be two real-valued functions defined on a topological space M
that are q-tame, i.e. such that the sublevel sets filtrations of f and g induce q-tame modules at
the homology level. Then for any integer k,

db (dgmk (f ), dgmk (g)) ≤ kf − gk∞ = sup |f (x) − g(x)|

x∈M

where dgmk (f ) (resp. dgmk (g)) is the persistence diagram of the persistence module (Hk (f −1 (−∞, r]))|r ∈
R) (resp. (Hk (g −1 (−∞, r]))|r ∈ R)) where the linear maps are the one induced by the canonical
inclusion maps between sublevel sets.

Proof. Denoting δ = kf − gk∞ we have that for any r ∈ R, f −1 (−∞, r]) ⊆ g −1 (−∞, r + δ])
and g −1 (−∞, r]) ⊆ f −1 (−∞, r + δ]). This interleaving between the sublevel sets of f induces a
δ-interleaving between the persistence modules at the homology level and the result follows from
the direct application of Theorem 7.

Theorem 7 also implies a stability result for the persistence diagrams of filtrations built on
top of data.

Theorem 9. Let X and Y be two compact metric spaces and let Filt(X) and Filt(Y) be the
Vietoris-Rips of Čech filtrations built on top X and Y. Then

db (dgm(Filt(X)), dgm(Filt(Y))) ≤ 2dGH (X, Y)

where dgm(Filt(X)) and dgm(Filt(Y)) denote the persistence diagram of the filtrations Filt(X)
and Filt(X).

As we already noticed in the Example 3 of Section 5.2, the persistence diagrams can be
interpreted as multiscale topological features of X and Y. In addition, Theorem 9 tells us that
these features are robust with respect to perturbations of the data in the Gromov-Hausdorff
metric. They can be used as discriminative features for classification or other tasks - see, for
example, Chazal et al. (2009b) for an application to non rigid 3D shapes classification.
We now give similar results for the alternative persistence homology representations intro-
duced before. From the definition of persistence landscape we immediately observe that λ(k, ·)
is one-Lipschitz and thus similar stability properties are satisfied for the landscapes as for per-
sistence diagrams.

Proposition 1 (Stability of persistence landscapes Bubenik (2015). ). Let dgm and dgm0 be
two persistence diagrams (without their essential parts). For any t ∈ R and any k ∈ N, we have:
(i) λ(k, t) ≥ λ(k + 1, t) ≥ 0.
(ii) |λ(k, t) − λ0 (k, t)| ≤ db (dgm, dgm0 )).

A large class of linear representations are continuous with respect to the Wasserstein metric
Ws in the space of persistence diagrams and with respect to the Banach norm of the linear
representation of persistence. Generally speaking, it is not always possible to upper bound the
modulus of continuity of the linear representation operator. However, in the case where s = 1 it
is even possible to show a stability result if the weight function take small values for points close
to the diagonal, see Divol and Lacombe (2020); Hofer et al. (2019b).

Stability versus discriminative capacity of persistence representations. The results

of Divol and Lacombe (2020) show that continuity and stability is only possible with weigh
functions taking small values for points close to the diagonal. However, in general there is no
specific reason to consider that points close to the diagonal are less important than others, given a
learning task. In a machine learning perspective, it is also relevant to design linear representation
with general weigh functions, although it would be more difficult to prove consistency of the

27
corresponding methods without at least the continuity of the representation. Stability is thus
important but maybe a too strong requirement for many problems in data sciences. Designing
linear representation that are sensitive to specific parts of a persistence diagrams rather than
globally stable may reveal a good strategy in practice.

6 Statistical aspects of persistent homology

Persistence homology by itself does not take into account the random nature of data and the
intrinsic variability of the topological quantity they infer. We now present a statistical approach
to persistent homology, in the sense that data is considered as generated from an unknown
distribution. We start with several consistency results for persistent homology inference.

6.1 Consistency results for persistent homology

Assume that we observe n points (X1 , . . . , Xn ) in a metric space (M, ρ) drawn i.i.d. from an
unknown probability measure µ whose support is a compact set denoted Xµ . The Gromov-
Hausdorff distance allows us to compare Xµ with compact metric spaces not necessarily embedded
in M . In the following, an estimator X b of Xµ is a function of X1 . . . , Xn that takes values in the
set of compact metric spaces.
Let Filt(Xµ ) and Filt(X)
b be two filtrations defined on Xµ and X. b Starting from Theorem 9;
an natural strategy for estimating the persistent homology of Filt(Xµ ) consists in estimating
the support Xµ . Note that in some cases the space M can be unknown and the observations
X1 . . . , Xn are then only known through their pairwise distances ρ(Xi , Xj ), i, j = 1, · · · , n.
The use of the Gromov-Hausdorff distance allows us to consider this set of observations as an
abstract metric space of cardinality n, independently of the way it is embedded in M . This
general framework includes the more standard approach consisting in estimating the support
with respect to the Hausdorff distance by restraining the values of X b to the compact sets included
in M .
The finite set Xn := {X1 , . . . , Xn } is a natural estimator of the support Xµ . In several
contexts discussed in the following, Xn shows optimal rates of convergence to Xµ with respect to
the Hausdorff distance. For some constants a, b > 0, we say that µ satisfies the (a, b)-standard
assumption if for any x ∈ Xµ and any r > 0,

µ(B(x, r)) ≥ min(arb , 1). (6)

This assumption has been widely used in the literature of set estimation under Hausdorff distance
(Cuevas and Rodríguez-Casal, 2004; Singh et al., 2009). Under this assumption, it can be easily
derived that the rate of convergence of dgm(Filt(Xn )) to dgm(Filt(Xµ )) for the bottleneck metric
 1/b
is upper bounded by O logn n . More precisely, this rate upper bounds the minimax rate of
convergence over the set of probability measures on the metric space (M, ρ) satisfying the (a, b)-
standard assumption on M .

Theorem 10. [Chazal et al. (2014b)] For some positive constants a and b, let
n  o
P := µ on M | Xµ is compact and ∀x ∈ Xµ , ∀r > 0, µ (B(x, r)) ≥ min 1, arb .

Then, it holds
 1/b
log n
sup E [db (dgm(Filt(Xµ )), dgm(Filt(Xn )))] ≤ C
µ∈P n
where the constant C only depends on a and b.

28
 Under additional technical assumptions, the corresponding lower bound can be shown (up to
a logarithmic term), see Chazal et al. (2014b). By applying stability results, similar consistency
results can be easily derived under alternative generative models as soon as a consistent estimator
of the support under Hausdorff metric is known. For instance, from the results of Genovese et al.
(2012) about Hausdorff support estimation under additive noise, it can be deduced that the
minimax convergence rates for the persistence diagram estimation is faster than (log n)−1/2 .
Moreover, as soon as a stability result is available for some given representation of persistence,
similar consistency results can be directly derived from the consistency for persistence diagrams.

Estimation of the persistent homology of functions. Theorem 7 opens the door to the
estimation of the persistent homology of functions defined on Rd , on a submanifold of Rd or
more generally on a metric space. The persistent homology of regression functions has also
been studied in Bubenik et al. (2010). The alternative approach of Bobrowski et al. (2014)
which is based on the inclusion map between nested pairs of estimated level sets can be applied
with kernel density and regression kernel estimators to estimate persistence homology of density
functions and regression functions. Another direction of research on this topic concerns various
versions of robust TDA. One solution is to study the persistent homology of the upper level
sets of density estimators (Fasy et al., 2014b). A different approach, more closely related to the
distance function, but robust to noise, consists in studying the persistent homology of the sub
level sets of the distance to measure defined in Section 4.4 (Chazal et al., 2014a).

6.2 Statistic of persistent homology computed on a point cloud

For many applications, in particular when the support of the point cloud is not drawn on or close
to a geometric shape, persistence diagrams can be quite complex to analyze. In particular, many
topological features are closed to the diagonal. Since they correspond to topological structures
that die very soon after they appear in the filtration, these points are generally considered as
noise, see Figure 13 for an illustration. Confidence regions of persistence diagram are rigorous
answers to the problem of distinguishing between signal and noise in these representations.
The stability results given in Section 5.7 motivate the use of the bottleneck distance to define
confidence regions. However alternative distances in the spirit of Wasserstein distances can be
proposed too. When estimating a persistence diagram dgm with an estimator dgm, d we typically
look for some value ηα such that
d dgm) ≥ ηα ) ≤ α,
P (db (dgm,

for α ∈ (0, 1). Let Bα be the closed ball of radius α for the bottleneck distance and centered at
d in the space of persistence diagrams. Following Fasy et al. (2014b), we can visualize the
dgm
signatures of the points belonging to this ball in various ways. One first option is to center a
box of side length 2α at each point of the persistence diagram dgm.
d An alternative solution is to
visualize the confidence set by adding a band at (vertical) distance ηα /2 from the diagonal (the
bottleneck distance being defined for the `∞ norm), see Figure 13 for an illustration. The points
outside the band are then considered as significant topological features, see Fasy et al. (2014b)
for more details.
Several methods have been proposed in Fasy et al. (2014b) to estimate ηα in different frame-
works. These methods mainly rely from stability results for persistence diagrams: confidence
sets for diagrams can be derived from confidence sets in the sample space.

Subsampling approach. This method is based on a confidence region for the support K of
the distribution of the sample in Hausdorff distance. Let X̃b be a subsample of size b drawn
from thesampleX̃n , where b = o(n/logn). Let qb (1 − α) be the quantile of of the distribution
of Haus X̃b , Xn . Take η̂α := 2q̂b (1 − α) where q̂b is an estimation qb (1 − α) using a standard

29
Monte Carlo procedure. Under an (a, b) standard assumption, and for n large enough, Fasy et al.
(2014b) show that


P (db (dgm (Filt(K)) , dgm (Filt(Xn ))) > η̂α ) ≤ P Haus (K, Xn ) > η̂α
 1/4
b
≤ α+O .
n

Bottleneck Bootstrap. The stability results often leads to conservative confidence sets. An
alternative strategy is the bottleneck bootstrap introduced in Chazal et al. (2016b). We consider
the general setting where a persistence diagram dgm d is defined from the observation (X1 , . . . , Xn )
in a metric space. This persistence diagram corresponds to the estimation of an underlying
persistence diagram dgm, which can be related for instance to the support of the measure, or to
the sublevel sets of a function related to this distribution (for instance a density function when
the Xi ’s are in Rd ). Let (X1∗ , . . . , Xn∗ ) be a sample from the empirical measure defined from the
observations (X1 , . . . , Xn ). Let also dgm d ∗ be the persistence diagram derived from this sample.
We then can take for ηα the quantity η̂α defined by

d ∗ , dgm)
P (db (dgm d > η̂α | X1 , . . . , Xn ) = α. (7)

Note that η̂α can be easily estimated with Monte Carlo procedures. It has been shown in Chazal
et al. (2016b) that the bottleneck bootstrap is valid when computing the sublevel sets of a density
estimator.

Bootstrapping Persistent Betti numbers. As already mentioned, confidence regions based

on stability properties of persistence may lead to very conservative confidence regions. Based
on the concepts of stabilizing statistics Penrose and Yukich (2001), asymptotic normality for
persistent Betti numbers has been shown recently by Krebs and Polonik (2019); Roycraft et al.
(2020), under very mild conditions on the filtration and the distribution of sample cloud. In
addition, bootstrap procedures are also shown to be valid in this framework. More precisely, a
smoothed bootstrap procedure together with a convenient rescaling of the point cloud seems to
be a promising approach for boostrapping TDA features from point cloud data.

6.3 Statistic for a family of persistent diagrams or other representations

Up to now in this section, we were only considering statistics based on one single observed
persistence diagram. We now consider a new framework where several persistence diagrams
(or other representations) are available and we are interested in providing central tendency,
confidence regions and hypothesis tests for topological descriptors built on this family.

6.3.1 Central tendency for persistent homology

Mean and expectations of distributions of diagrams. The space of persistence diagrams
being a general metric space but not an Hilbert space, the definition of a mean persistence
diagram is not obvious and unique. One first natural approach to define a central tendency in
this context is to consider Fréchet means of distributions of diagrams. Their existence has been
proved in Mileyko et al. (2011) and they have also been characterized in Turner et al. (2014a).
However they are may not be unique and they turn out to be difficult to compute in practice.
To partly overcome these problems, different approaches have been recently proposed based on
numerical optimal transport Lacombe et al. (2018) or linear representations and kernel based
methods Divol and Chazal (2020).

30
Topological signatures from subsamples Central tendency properties of persistent homol-
ogy can also be used to compute topological signatures for very large data sets, as an alternative
approach to overcome the prohibitive cost of persistence computations. Given a large point
cloud, the idea is to extract many subsamples, to compute the persistence landscape for each
subsample and then to combine the information.
For any positive integer m, let X = {x1 , · · · , xm } be a sample of m points drawn from
a measure µ in a metric space M and which support is denoted Xµ . We assume that the
diameter of Xµ is finite and upper bounded by T2 , where T is the same constant as in the
definition of persistence landscapes in Section 5.4. For ease of exposition, we focus on the case
k = 1, and set λ(t) = λ(1, t). However, the results we present in this section hold for k > 1.
The corresponding persistence landscape (associated to the persistence diagram of the Čech
or Rips-Vietoris filtration) is λX and we denote by Ψm µ the measure induced by µ
⊗m on the

space of persistence landscapes. Note that the persistence landscape λX can be seen as a single
draw from the measure Ψm µ . The point-wise expectations of the (random) persistence landscape
under this measure is defined by EΨm µ
[λX (t)], t ∈ [0, T ]. The average landscape EΨm
µ
[λX ] has a
natural empirical counterpart, which can be used as its unbiased estimator. Let S1 , . . . , S`m be
m

` independent samples of size m from µ⊗m . We define the empirical average landscape as
b
1X
λm
` (t) = λSim (t), for all t ∈ [0, T ], (8)
b
i=1

and propose to use λm ` to estimate λXµ . Note that computing the persistent homology of Xn is
O(exp(n)), whereas computing the average landscape is O(b exp(m)).
Another motivation for this subsampling approach is that it can be also applied when µ is a
discrete measure with support XN = {x1 , . . . , xN } lying in a metric space M . This framework
can be very common in practice, when a continuous (but unknown measure) is approximated by
a discrete uniform measure µN on XN .
The average landscape EΨm µ
[λX ] is an interesting quantity on its own, since it carries some
stable topological information about the underlying measure µ, from which the data are gener-
ated.

Theorem 11. [Chazal et al. (2015a)] Let X ∼ µ⊗m and Y ∼ ν ⊗m , where µ and ν are two
probability measures on M . For any p ≥ 1 we have
1
EΨµ [λX ] − EΨm [λ ] ≤ 2 m p Wp (µ, ν),
m
ν Y
∞

where Wp is the pth Wasserstein distance on M .

The result of Theorem 11 is useful for two reasons. First, it tells us that for a fixed m, the
expected "topological behavior" of a set of m points carries some stable information about the
underlying measure from which the data are generated. Second, it provides a lower bound for
the Wasserstein distance between two measures, based on the topological signature of samples
of m points.

6.3.2 Asymptotic normality

As in the previous section, we consider several persistence diagrams (or other representations).
The next step after giving central tendency descriptors of persistence homology is to provide
asymptotic normality results for these quantities together with bootstrap procedures to derive
confidence regions. It is of course easier to show such results for functional representations
of persistence. In Chazal et al. (2013a, 2015b), following this strategy confidence bands for
landscapes are proposed from the observation of landscapes λ1 , . . . , λN drawn i.i.d. from a
random distribution in the space of landscapes. The asymptotic validity as well as the uniform

31
convergence of the multiplier bootstrap is shown in this framework. Note that similar results
can be also proposed for many representations of persistence, in particular by showing that the
corresponding functional spaces are Donsker spaces.

6.4 Other statistical approaches to TDA

Statistical approaches for tda are knowing an increasing interest and many others have been
proposed in the recent years or are still subject to active research activities, as illustrated in the
following non-exhaustive list of examples.

Hypothesis testing. Several methods have been proposed for hypothesis testing procedures
for persistent homology, mostly based on permutation strategies and for two sample testing.
Robinson and Turner (2017) focuses on pairwise distances of persistence diagrams whereas Berry
et al. (2020) study more general functional summaries. Hypothesis tests based on kernel ap-
proaches have been proposed in Kusano (2019). A two-stage hypothesis test of filtering and
testing for persistent images is also presented inMoon and Lazar (2020).

Persistence Homology Transform. The representations introduced before are all transfor-
mations derived from the persistence diagram computed from a fixed filtration built over a data
set. The Persistence Homology Transform introduced in Curry et al. (2018); Turner et al. (2014b)
to study shapes in Rd , takes a different path by looking at the persistence homology of the sub-
level set filtration induced by the projection of the considered shape to each direction in Rd . It
comes with several interesting properties, in particular the Persistence Homology Transform is a
sufficient statistic for distributions defined on the set of geometric and finite simplicial complexes
embedded in Rd .

Bayesian statistics for TDA. A Bayesian approach to persistence diagram inference has
been proposed in Maroulas et al. (2020) by viewing a persistence diagram as a sample from a
point process. This Bayesian method computes the point process posterior intensity based on a
Gaussian mixture intensity for the prior.

6.5 Persistent homology and machine learning

Using tda and more specifically persistent homology for machine learning is a subject that
attracts a lot of information and generated an intense research activity. Although the recent
progress in this area goes far beyond the scope of this paper, we briefly introduce the main
research directions with a few reference to help the newcomer to the field to get started.

tda for exploratory data analysis and descriptive statistics. In some domains, tda can
be fruitfully used as a tool for exploratory analysis and visualization. For example, the Mapper
algorithm provides a powerful approach to explore and vizualize the global topological structure
of complex data sets. In some cases, persistence diagrams obtained from data can be directly
interpreted and exploited for better understanding of the phenomena from which the data have
been generated. This is, for example, the case in the study of force fields in granular media
(Kramar et al., 2013) or of atomic structures in glass (Nakamura et al., 2015) in material science,
in the study of the evolution of convection patterns in fluid dynamics (Kramár et al., 2016), in
machining monitoring (Khasawneh and Munch, 2016) or in the analysis of nanoporous structures
in chemistry (Lee et al., 2017) where topological features can be rather clearly related to specific
geometric structures and patterns in the considered data.

32
Persistent homology for feature engineering. There are many other cases where persis-
tence features cannot be easily or directly interpreted but present valuable information for further
processing. However, the highly non linear nature of diagrams prevents them to be immediately
used as standard features in machine learning algorithms.
Persistence landscapes and linear representations of persistence diagrams offer a first option
to convert persistence diagrams into elements of a vector space that can be directly used as
features in classical machine learning pipelines. This approach has been used, for example, for
protein binding (Kovacev-Nikolic et al., 2016), object recognition (Li et al., 2014) or time-series
analysis. In the same vein, the construction of kernels for persistence diagrams that preserve
their stability properties has recently attracted some attention. Most of them have been obtained
by considering diagrams as discrete measures in R2 . Convolving a symmetrized (with respect to
the diagonal) version of persistence diagrams with a 2D Gaussian distribution, Reininghaus et al.
(2015) introduce a multi-scale kernel and apply it to shape classification and texture recognition
problems. Considering Wasserstein distance between projections of persistence diagrams on lines,
Carriere and Oudot (2017) build another kernel and test its performance on several benchmarks.
Other kernels, still obtained by considering persistence diagrams as measures, have also been
proposed in Kusano et al. (2017).
Various other vector summaries of persistence diagrams have been proposed and then used
as features for different problems. For example, basic summaries are considered in Bonis et al.
(2016) and combined with quantization and pooling methods to address non rigid shape analysis
problems; Betti curves extracted from persistence diagrams are used with 1-dimensional Convo-
lutional Neural Networks (CNN) to analyze time dependent data and recognize human activities
from inertial sensors in Dindin et al. (2020); Umeda (2017); persistence images are introduced in
Adams et al. (2017) and are considered to address some inverse problems using linear machine
learning models in Obayashi and Hiraoka (2017).
The above mentioned kernels and vector summaries of persistence diagrams are built inde-
pendently of the considered data analysis or learning task. Moreover, it appears that in many
cases the relevant topological information is not carried by the whole persistence diagrams but
is concentrated in some localized regions that may not be obvious to identify. This usually
makes the choice of a relevant kernel or vector summary very difficult for the user. To overcome
this issue, various authors have proposed learning approaches that allows to learn the relevant
topological features for a given task. In this direction, Hofer et al. (2017) proposes a deep learn-
ing approach to learn parameters of persistence images representations of persistence diagrams
while Kim et al. (2020) introduce a neural network layer for persistence landscapes. In Carrière
et al. (2020), the authors introduce a general neural network layer for persistence diagrams that
can be either used to learn an appropriate vectorization or directly integrated in a deep neural
network architecture. Other methods, inspired from k-means, propose unsupervised methods to
vectorize persistence diagrams Royer et al. (2021); Zieliński et al. (2010), some of them coming
with theoretical guarantees Chazal et al. (2020).

Persistent homology for machine learning architecture optimization and model se-
lection. More recently, tda has found new developments in machine learning where persistent
homology is no longer used for feature engineering but as a tool to design, improve or select
models - see, e.g. Carlsson and Gabrielsson (2020); Chen et al. (2019); Gabrielsson and Carlsson
(2019); Hofer et al. (2019a); Moor et al. (2020); Ramamurthy et al. (2019); Rieck et al. (2019).
Many of these tools rely on the introduction of loss or regularization functions depending on per-
sistent homology features, raising the problem of their optimization. Building on the powerful
tools provided by software libraries such as PyTorch or TensorFlow, practical methods allowing
to encode and optimize a large family of persistence-based functions have been proposed and
experimented Brüel-Gabrielsson et al. (2019); Poulenard et al. (2018). A general framework for
persistence-based function optimization based on stochastic subgradient descent algorithms with

33
 convergence guarantees has been recently proposed and implemented in a easy-to-use software
tool Carriere et al. (2020). With a different perspective, another theoretical framework to study
the differentiable structure of functions of persistence diagrams has been proposed in Leygonie
et al. (2019).

7 tda for data sciences with the GUDHI library

In this section we illustrate TDA methods with the Python library Gudhi11 (Maria et al., 2014)
together with popular libraries as numpy (Walt et al., 2011), scikit-learn (Pedregosa et al., 2011),
pandas (McKinney et al., 2010). More illustrations with python notebooks can be found in the
Tutorial Github page12 of Gudhi.

7.1 Bootstrap and comparison of protein binding configurations

This example is borrowed from Kovacev-Nikolic et al. (2016). In this paper, persistent homology
is used to analyze protein binding and more precisely it compares closed and open forms of the
maltose-binding protein (MBP), a large biomolecule consisting of 370 amino acid residues. The
analysis is not based on geometric distances in R3 but on a metric of dynamical distances defined
by
Dij = 1 − |Cij |,
where C is the correlation matrices between residues. The data can be download at this link13 .
1 import numpy a s np
2 import gudhi a s gd
3 import pandas a s pd
4 import seaborn as sns
5
6 c o r r _ p r o t e i n = pd . read_csv ( "mypath /1 a n f . corr_1 . t x t " ,
7 h e a d e r=None ,
8 d e l i m _ w h i t e s p a c e=True )
9 d i s t _ p r o t e i n _ 1 = 1− np . abs ( c o r r _ p r o t e i n _ 1 . v a l u e s )
10 rips_complex_1= gd . RipsComplex ( d i s t a n c e _ m a t r i x=d i s t _ p r o t e i n _ 1 ,
11 max_edge_length =1.1)
12 simplex_tree_1 = rips_complex_1 . c r e a t e _ s i m p l e x _ t r e e ( max_dimension=2)
13 diag_1 = simplex_tree_1 . p e r s i s t e n c e ( )
14 gd . p l o t _ p e r s i s t e n c e _ d i a g r a m ( diag_1 )
For comparing persistence diagrams, we use the bottleneck distance. The block of statements
given below computes persistence intervals and computes the bottleneck distance for 0-homology
and 1-homology:
1 i n t e r v 0 _ 1 = simplex_tree_1 . p e r s i s t e n c e _ i n t e r v a l s _ i n _ d i m e n s i o n ( 0 )
2 i n t e r v 0 _ 2 = simplex_tree_2 . p e r s i s t e n c e _ i n t e r v a l s _ i n _ d i m e n s i o n ( 0 )
3 bot0 = gd . b o t t l e n e c k _ d i s t a n c e ( interv0_1 , i n t e r v 0 _ 2 )
4
5 i n t e r v 1 _ 1 = simplex_tree_1 . p e r s i s t e n c e _ i n t e r v a l s _ i n _ d i m e n s i o n ( 1 )
6 i n t e r v 1 _ 2 = simplex_tree_2 . p e r s i s t e n c e _ i n t e r v a l s _ i n _ d i m e n s i o n ( 1 )
7 bot1 = gd . b o t t l e n e c k _ d i s t a n c e ( interv1_1 , i n t e r v 1 _ 2 )
In this way, we can compute the matrix of bottleneck distances between the fourteen MPB.
Finally, we apply a multidimensional scaling method to find a configuration in R2 which almost
match with the bottleneck distances, see Figure 13(c). We use the scikit-learn library for the
MDS:
1 import m a t p l o t l i b . p y p l o t a s p l t
2 from s k l e a r n import m a n i f o l d
11
http://gudhi.gforge.inria.fr/python/latest/
12
https://github.com/GUDHI/TDA-tutorial
13
https://www.researchgate.net/publication/301543862_corr

34
 Figure 13: (a) and (b) two persistence diagrams for two configurations of MBP. (c) MDS con-
figuration for the matrix of bottleneck distances. (d) Persistence diagram and confidence region
for the persistence diagram of a MBP.

3
4 mds = m a n i f o l d .MDS( n_components =2, d i s s i m i l a r i t y=" precomputed " )
5 c o n f i g = mds . f i t (M) . embedding_
6
7 p l t . s c a t t e r ( c o n f i g [ 0 : 7 , 0 ] , c o n f i g [ 0 : 7 , 1 ] , c o l o r= ’ r e d ’ , l a b e l=" c l o s e d " )
8 p l t . s c a t t e r ( c o n f i g [ 7 : l , 0 ] , c o n f i g [ 7 : l , 1 ] , c o l o r= ’ b l u e ’ , l a b e l=" r e d " )
9 p l t . l e g e n d ( l o c =1)

We now define a confidence band for a diagram using the bottleneck bootstrap approach. We
resample over the lines (and columns) of the matrix of distances and we compute the bottleneck
distance between the original persistence diagram and the bootstrapped persistence diagram.
We repeat the procedure many times and finally we estimate the quantile 95% of this collection
of bottleneck distances. We take the value of the quantile to define a confidence band on the
original diagram (see Figure 13(d)). However, such a procedure should be considered with caution
because as far as we know the validity of the bottleneck bootstrap has not been proved in this
framework.

35
 7.2 Classification for sensor data
In this experiment, the 3d acceleration of 3 walkers (A, B and C) have been recorded from the
sensor of a smart phone14 . Persistence homology is not sensitive to the choice of axes and so
no preprocessing is necessary to align the 3 times series according to the same axis. From these
three times series, we have picked at random sequences of 8 seconds in the complete time series,
that is 200 consecutive points of acceleration in R3 . For each walker, we extract 100 time series in
this way. The next block of statements computes the persistence for the alpha complex filtration
for data_A_sample, one of the 100 times series of acceleration of Walker A.
1 alpha_complex_sample = gd . AlphaComplex ( p o i n t s = data_A_sample )
2 simplex_tree_sample = alpha_complex_sample . c r e a t e _ s i m p l e x _ t r e e ( max_alpha_square
=0.3)
3 diag_Alpha = simplex_tree_sample . p e r s i s t e n c e ( )

From diag_Alpha we can then easily compute and plot the persistence landscapes, see Fig-
ure 14(a). For all the 300 times series, we compute the persistence landscapes for dimension 0
and 1 and we compute the three first landscapes for the 2 dimensions. Moreover, each persistence
landscape is discretized on 1000 points. Each time series is thus described by 6000 topological
variables. To predict the walker from these features, we use a random forest (Breiman, 2001),
which is known to be an efficient in such an high dimensional setting. We split the data into
train and test samples at random several times. We finally obtain an averaged classification
error around 0.95. We can also visualize the most important variables in the Random Forest,
see Figure 14(b).

Figure 14: (a)The three first landscapes for 0-homology of the alpha shape filtration defined for
a time series of acceleration of Walker A. (b) Variable importances of the landscapes coefficients
for the classification of Walkers. The 3 000 first coefficients correspond to the three landscapes of
dimension 0 and the 3 000 last coefficients to the three landscapes of dimension 1. There are 1000
coefficients per landscape. Note that the first landscape of dimension 0 is always the same using
the Rips complex (a trivial landscape) and consequently the corresponding coefficients have a
zero importance value.

Acknowledgements This work was partly supported by the French ANR chair in Artificial
Intelligence TopAI. We thank the authors of Kovacev-Nikolic et al. (2016) for making their data
available.
14
The dataset can be download at this link http://bertrand.michel.perso.math.cnrs.fr/Enseignements/
TDA/data_acc

36
References
Aamari, E., Kim, J., Chazal, F., Michel, B., Rinaldo, A., Wasserman, L., et al. (2019). Estimating
the reach of a manifold. Electronic journal of statistics, 13(1):1359–1399.

Adams, H., Emerson, T., Kirby, M., Neville, R., Peterson, C., Shipman, P., Chepushtanova,
S., Hanson, E., Motta, F., and Ziegelmeier, L. (2017). Persistence images: a stable vector
representation of persistent homology. Journal of Machine Learning Research, 18(8):1–35.

Anai, H., Chazal, F., Glisse, M., Ike, Y., Inakoshi, H., Tinarrage, R., and Umeda, Y. (2020).
Dtm-based filtrations. In Topological Data Analysis, pages 33–66. Springer.

Balakrishna, S., Rinaldo, A., Sheehy, D., Singh, A., and Wasserman, L. A. (2012). Minimax rates
for homology inference. Journal of Machine Learning Research - Proceedings Track, 22:64–72.

Barannikov, S. (1994). The framed morse complex and its invariants. In Adv. Soviet Math.,
volume 21, pages 93–115. Amer. Math. Soc., Providence, RI.

Berry, E., Chen, Y.-C., Cisewski-Kehe, J., and Fasy, B. T. (2020). Functional summaries of
persistence diagrams. Journal of Applied and Computational Topology, 4(2):211–262.

Biau, G., Chazal, F., Cohen-Steiner, D., Devroye, L., and Rodriguez, C. (2011). A weighted
k-nearest neighbor density estimate for geometric inference. Electronic Journal of Statistics,
5:204–237.

Biscio, C. A. and Møller, J. (2019). The accumulated persistence function, a new useful functional
summary statistic for topological data analysis, with a view to brain artery trees and spatial
point process applications. Journal of Computational and Graphical Statistics, pages 1–21.

Bobrowski, O., Mukherjee, S., and Taylor, J. (2014). Topological consistency via kernel estima-
tion. arXiv preprint arXiv:1407.5272.

Bonis, T., Ovsjanikov, M., Oudot, S., and Chazal, F. (2016). Persistence-based pooling for shape
pose recognition. In Computational Topology in Image Context - 6th International Workshop,
CTIC 2016, Marseille, France, June 15-17, 2016, Proceedings, pages 19–29.

Brécheteau, C. et al. (2019). A statistical test of isomorphism between metric-measure spaces

using the distance-to-a-measure signature. Electronic journal of statistics, 13(1):795–849.

Brécheteau, C., Levrard, C., et al. (2020). A k-points-based distance for robust geometric infer-
ence. Bernoulli, 26(4):3017–3050.

Breiman, L. (2001). Random forests. Machine learning, 45(1):5–32.

Brown, A., Bobrowski, O., Munch, E., and Wang, B. (2020). Probabilistic convergence and
stability of random mapper graphs. Journal of Applied and Computational Topology, pages
1–42.

Brüel-Gabrielsson, R., Nelson, B. J., Dwaraknath, A., Skraba, P., Guibas, L. J., and Carlsson,
G. (2019). A topology layer for machine learning. arXiv preprint arXiv:1905.12200.

Bubenik, P. (2015). Statistical topological data analysis using persistence landscapes. Journal
of Machine Learning Research, 16:77–102.

Bubenik, P., Carlsson, G., Kim, P. T., and Luo, Z.-M. (2010). Statistical topology via morse the-
ory persistence and nonparametric estimation. Algebraic methods in statistics and probability
II, 516:75–92.

37
Buchet, M., Chazal, F., Dey, T. K., Fan, F., Oudot, S. Y., and Wang, Y. (2015a). Topological
analysis of scalar fields with outliers. In Proc. Sympos. on Computational Geometry.

Buchet, M., Chazal, F., Oudot, S., and Sheehy, D. R. (2015b). Efficient and robust persis-
tent homology for measures. In Proceedings of the 26th ACM-SIAM symposium on Discrete
algorithms. SIAM. SIAM.

Cadre, B. (2006). Kernel estimation of density level sets. Journal of multivariate analysis,
97(4):999–1023.

Carlsson, G. (2009). Topology and data. AMS Bulletin, 46(2):255–308.

Carlsson, G. and Gabrielsson, R. B. (2020). Topological approaches to deep learning. In Topo-

logical Data Analysis, pages 119–146. Springer.

Carriere, M., Chazal, F., Glisse, M., Ike, Y., and Kannan, H. (2020). A note on stochastic
subgradient descent for persistence-based functionals: convergence and practical aspects. arXiv
preprint arXiv:2010.08356.

Carrière, M., Chazal, F., Ike, Y., Lacombe, T., Royer, M., and Umeda, Y. (2020). Perslay:
a neural network layer for persistence diagrams and new graph topological signatures. In
International Conference on Artificial Intelligence and Statistics, pages 2786–2796. PMLR.

Carrière, M. and Michel, B. (2019). Approximation of reeb spaces with mappers and applications
to stochastic filters. arXiv preprint arXiv:1912.10742.

Carriere, M., Michel, B., and Oudot, S. (2018). Statistical analysis and parameter selection for
mapper. Journal of Machine Learning Research, 19(12).

Carrière, M. and Oudot, S. (2015). Structure and stability of the 1-dimensional mapper. arXiv
preprint arXiv:1511.05823.

Carriere, M. and Oudot, S. (2017). Sliced wasserstein kernel for persistence diagrams. To appear
in ICML-17.

Carrière, M. and Rabadán, R. (2020). Topological data analysis of single-cell hi-c contact maps.
In Topological Data Analysis, pages 147–162. Springer.

Chazal, F. (2017). High-dimensional topological data analysis. In Handbook of Discrete and

Computational Geometry (3rd Ed - To appear), chapter 27. CRC Press.

Chazal, F., Chen, D., Guibas, L., Jiang, X., and Sommer, C. (2011a). Data-driven trajectory
smoothing. In Proc. ACM SIGSPATIAL GIS.

Chazal, F., Cohen-Steiner, D., Glisse, M., Guibas, L., and Oudot, S. (2009a). Proximity of
persistence modules and their diagrams. In SCG, pages 237–246.

Chazal, F., Cohen-Steiner, D., Guibas, L. J., M’emoli, F., and Oudot, S. Y. (2009b). Gromov-
hausdorff stable signatures for shapes using persistence. Computer Graphics Forum (proc.
SGP 2009), pages 1393–1403.

Chazal, F., Cohen-Steiner, D., and Lieutier, A. (2009c). Normal cone approximation and offset
shape isotopy. Comp. Geom. Theor. Appl., 42(6-7):566–581.

Chazal, F., Cohen-Steiner, D., and Lieutier, A. (2009d). A sampling theory for compact sets in
euclidean space. Discrete & Computational Geometry, 41(3):461–479.

38
Chazal, F., Cohen-Steiner, D., Lieutier, A., and Thibert, B. (2008). Stability of Curvature
Measures. Computer Graphics Forum (proc. SGP 2009), pages 1485–1496.

Chazal, F., Cohen-Steiner, D., and Mérigot, Q. (2010). Boundary measures for geometric infer-
ence. Found. Comp. Math., 10:221–240.

Chazal, F., Cohen-Steiner, D., and Mérigot, Q. (2011b). Geometric inference for probability
measures. Foundations of Computational Mathematics, 11(6):733–751.

Chazal, F., de Silva, V., Glisse, M., and Oudot, S. (2016a). The structure and stability of
persistence modules. SpringerBriefs in Mathematics. Springer.

Chazal, F., Fasy, B. T., Lecci, F., Michel, B., Rinaldo, A., and Wasserman, L. (2014a). Robust
topological inference: Distance to a measure and kernel distance. to appear in JMLR.

Chazal, F., Fasy, B. T., Lecci, F., Michel, B., Rinaldo, A., and Wasserman, L. (2015a). Subsam-
pling methods for persistent homology. To appear in Proceedings of the 32 st International
Conference on Machine Learning (ICML-15).

Chazal, F., Fasy, B. T., Lecci, F., Rinaldo, A., Singh, A., and Wasserman, L. (2013a). On the
bootstrap for persistence diagrams and landscapes. arXiv preprint arXiv:1311.0376.

Chazal, F., Fasy, B. T., Lecci, F., Rinaldo, A., and Wasserman, L. (2015b). Stochastic conver-
gence of persistence landscapes and silhouettes. Journal of Computational Geometry, 6(2):140–
161.

Chazal, F., Glisse, M., Labruère, C., and Michel, B. (2014b). Convergence rates for persistence
diagram estimation in topological data analysis. To appear in Journal of Machine Learning
Research.

Chazal, F., Guibas, L. J., Oudot, S. Y., and Skraba, P. (2013b). Persistence-based clustering in
riemannian manifolds. Journal of the ACM (JACM), 60(6):41.

Chazal, F., Huang, R., and Sun, J. (2015c). Gromov—hausdorff approximation of filamentary
structures using reeb-type graphs. Discrete Comput. Geom., 53(3):621–649.

Chazal, F., Levrard, C., and Royer, M. (2020). Optimal quantization of the mean measure and
application to clustering of measures. arXiv preprint arXiv:2002.01216.

Chazal, F. and Lieutier, A. (2008a). Smooth manifold reconstruction from noisy and non-uniform
approximation with guarantees. Comp. Geom. Theor. Appl., 40(2):156–170.

Chazal, F. and Lieutier, A. (2008b). Smooth manifold reconstruction from noisy and non uniform
approximation with guarantees. Computational Geometry Theory and Applications, 40:156–
170.

Chazal, F., Massart, P., and Michel, B. (2016b). Rates of convergence for robust geometric
inference. Electron. J. Statist, 10:2243–2286.

Chazal, F. and Oudot, S. Y. (2008). Towards persistence-based reconstruction in euclidean

spaces. In Proceedings of the twenty-fourth annual symposium on Computational geometry,
SCG ’08, pages 232–241, New York, NY, USA. ACM.

Chen, C., Ni, X., Bai, Q., and Wang, Y. (2019). A topological regularizer for classifiers via persis-
tent homology. In The 22nd International Conference on Artificial Intelligence and Statistics,
pages 2573–2582.

39
Chen, Y.-C., Genovese, C. R., and Wasserman, L. (2015). Density level sets: Asymptotics,
inference, and visualization. arXiv preprint arXiv:1504.05438.

Chen, Y.-C., Genovese, C. R., and Wasserman, L. (2017). Density level sets: Asymptotics,
inference, and visualization. Journal of the American Statistical Association, 112(520):1684–
1696.

Cohen-Steiner, D., Edelsbrunner, H., and Harer, J. (2005). Stability of persistence diagrams. In
SCG, pages 263–271.

Cohen-Steiner, D., Edelsbrunner, H., Harer, J., and Mileyko, Y. (2010). Lipschitz functions have
l p-stable persistence. Foundations of computational mathematics, 10(2):127–139.

Cuevas, A. and Rodríguez-Casal, A. (2004). On boundary estimation. Adv. in Appl. Probab.,

36(2):340–354.

Cuevas, A. and Rodríguez-Casal, A. (2004). On boundary estimation. Advances in Applied

Probability, pages 340–354.

Curry, J., Mukherjee, S., and Turner, K. (2018). How many directions determine a shape and
other sufficiency results for two topological transforms. arXiv preprint arXiv:1805.09782.

De Silva, V. and Carlsson, G. (2004). Topological estimation using witness complexes. In

Proceedings of the First Eurographics Conference on Point-Based Graphics, SPBG’04, pages
157–166, Aire-la-Ville, Switzerland, Switzerland. Eurographics Association.

De Silva, V. and Ghrist, R. (2007). Homological sensor networks. Notices of the American
mathematical society, 54(1).

Devroye, L. and Wise, G. L. (1980). Detection of abnormal behavior via nonparametric estima-
tion of the support. SIAM J. Appl. Math., 38(3):480–488.

Dey, T. K., Mémoli, F., and Wang, Y. (2016). Multiscale mapper: topological summarization
via codomain covers. In Proceedings of the Twenty-Seventh Annual ACM-SIAM Symposium
on Discrete Algorithms, pages 997–1013. Society for Industrial and Applied Mathematics.

Dey, T. K., Mémoli, F., and Wang, Y. (2017). Topological analysis of nerves, reeb spaces,
mappers, and multiscale mappers. In Proc. Sympos. Comput. Geom. (SoCG).

Dindin, M., Umeda, Y., and Chazal, F. (2020). Topological data analysis for arrhythmia detection
through modular neural networks. In Canadian Conference on Artificial Intelligence, pages
177–188. Springer.

Divol, V. and Chazal, F. (2020). The density of expected persistence diagrams and its kernel
based estimation. Journal of Computational Geometry, 10(2):127–153.

Divol, V. and Lacombe, T. (2020). Understanding the topology and the geometry of the persis-
tence diagram space via optimal partial transport. J Appl. and Comput. Topology.

Edelsbrunner, H., Letscher, D., and Zomorodian, A. (2002). Topological persistence and simpli-
fication. Discrete Comput. Geom., 28:511–533.

Fasy, B. T., Kim, J., Lecci, F., and Maria, C. (2014a). Introduction to the r package tda. arXiv
preprint arXiv:1411.1830.

Fasy, B. T., Lecci, F., Rinaldo, A., Wasserman, L., Balakrishnan, S., and Singh, A. (2014b).
Confidence sets for persistence diagrams. The Annals of Statistics, 42(6):2301–2339.

40
Federer, H. (1959). Curvature measures. Transactions of the American Mathematical Society,
pages 418–491.

Frosini, P. (1992). Measuring shapes by size functions. In Intelligent Robots and Computer
Vision X: Algorithms and Techniques, volume 1607, pages 122–133. International Society for
Optics and Photonics.

Gabrielsson, R. B. and Carlsson, G. (2019). Exposition and interpretation of the topology of

neural networks. In 2019 18th IEEE International Conference On Machine Learning And
Applications (ICMLA), pages 1069–1076. IEEE.

Genovese, C. R., Perone-Pacifico, M., Verdinelli, I., and Wasserman, L. (2012). Manifold esti-
mation and singular deconvolution under hausdorff loss. Ann. Statist., 40:941–963.

Ghrist, R. (2017). Homological algebra and data. preprint.

Grove, K. (1993). Critical point theory for distance functions. In Proc. of Symposia in Pure
Mathematics, volume 54.

Guibas, L., Morozov, D., and Mérigot, Q. (2013). Witnessed k-distance. Discrete Comput.
Geom., 49:22–45.

Hatcher, A. (2001). Algebraic Topology. Cambridge Univ. Press.

Hofer, C., Kwitt, R., Niethammer, M., and Dixit, M. (2019a). Connectivity-optimized represen-
tation learning via persistent homology. In Proceedings of the 36th International Conference on
Machine Learning, volume 97 of Proceedings of Machine Learning Research, pages 2751–2760.
PMLR.

Hofer, C., Kwitt, R., Niethammer, M., and Uhl, A. (2017). Deep learning with topological
signatures. arXiv preprint arXiv:1707.04041.

Hofer, C. D., Kwitt, R., and Niethammer, M. (2019b). Learning representations of persistence
barcodes. Journal of Machine Learning Research, 20(126):1–45.

Khasawneh, F. A. and Munch, E. (2016). Chatter detection in turning using persistent homology.
Mechanical Systems and Signal Processing, 70:527–541.

Kim, K., Kim, J., Zaheer, M., Kim, J., Chazal, F., and Wasserman, L. (2020). Pllay: Efficient
topological layer based on persistence landscapes. In 34th Conference on Neural Information
Processing Systems (NeurIPS 2020).

Kovacev-Nikolic, V., Bubenik, P., Nikolić, D., and Heo, G. (2016). Using persistent homology
and dynamical distances to analyze protein binding. Statistical applications in genetics and
molecular biology, 15(1):19–38.

Kramar, M., Goullet, A., Kondic, L., and Mischaikow, K. (2013). Persistence of force networks
in compressed granular media. Physical Review E, 87(4):042207.

Kramár, M., Levanger, R., Tithof, J., Suri, B., Xu, M., Paul, M., Schatz, M. F., and Mischaikow,
K. (2016). Analysis of kolmogorov flow and rayleigh–bénard convection using persistent ho-
mology. Physica D: Nonlinear Phenomena, 334:82–98.

Krebs, J. T. and Polonik, W. (2019). On the asymptotic normality of persistent betti numbers.
arXiv preprint arXiv:1903.03280.

Kusano, G. (2019). On the expectation of a persistence diagram by the persistence weighted

kernel. Japan Journal of Industrial and Applied Mathematics, 36(3):861–892.

41
Kusano, G., Fukumizu, K., and Hiraoka, Y. (2017). Kernel method for persistence diagrams via
kernel embedding and weight factor. arXiv preprint arXiv:1706.03472.

Kusano, G., Hiraoka, Y., and Fukumizu, K. (2016). Persistence weighted gaussian kernel for
topological data analysis. In International Conference on Machine Learning, pages 2004–2013.

Lacombe, T., Cuturi, M., and Oudot, S. (2018). Large scale computation of means and clusters
for persistence diagrams using optimal transport. In NeurIPS.

Lee, Y., Barthel, S. D., Dłotko, P., Moosavi, S. M., Hess, K., and Smit, B. (2017). Quantifying
similarity of pore-geometry in nanoporous materials. Nature Communications, 8.

Leygonie, J., Oudot, S., and Tillmann, U. (2019). A framework for differential calculus on
persistence barcodes. arXiv preprint arXiv:1910.00960.

Li, C., Ovsjanikov, M., and Chazal, F. (2014). Persistence-based structural recognition. In
Computer Vision and Pattern Recognition (CVPR), 2014 IEEE Conference on, pages 2003–
2010.

Lum, P., Singh, G., Lehman, A., Ishkanov, T., Vejdemo-Johansson, M., Alagappan, M., Carlsson,
J., and Carlsson, G. (2013). Extracting insights from the shape of complex data using topology.
Scientific reports, 3.

Maria, C., Boissonnat, J.-D., Glisse, M., and Yvinec, M. (2014). The gudhi library: Simplicial
complexes and persistent homology. In International Congress on Mathematical Software,
pages 167–174. Springer.

Maroulas, V., Nasrin, F., and Oballe, C. (2020). A bayesian framework for persistent homology.
SIAM Journal on Mathematics of Data Science, 2(1):48–74.

McKinney, W. et al. (2010). Data structures for statistical computing in python. In Proceedings
of the 9th Python in Science Conference, volume 445, pages 51–56. SciPy Austin, TX.

Mileyko, Y., Mukherjee, S., and Harer, J. (2011). Probability measures on the space of persistence
diagrams. Inverse Problems, 27(12):124007.

Moon, C. and Lazar, N. A. (2020). Hypothesis testing for shapes using vectorized persistence
diagrams. arXiv preprint arXiv:2006.05466.

Moor, M., Horn, M., Rieck, B., and Borgwardt, K. (2020). Topological autoencoders. In Inter-
national Conference on Machine Learning, pages 7045–7054. PMLR.

Nakamura, T., Hiraoka, Y., Hirata, A., Escolar, E. G., and Nishiura, Y. (2015). Persistent ho-
mology and many-body atomic structure for medium-range order in the glass. Nanotechnology,
26(30):304001.

Niyogi, P., Smale, S., and Weinberger, S. (2008). Finding the homology of submanifolds with
high confidence from random samples. Discrete & Computational Geometry, 39(1-3):419–441.

Niyogi, P., Smale, S., and Weinberger, S. (2011). A topological view of unsupervised learning
from noisy data. SIAM Journal on Computing, 40(3):646–663.

Obayashi, I. and Hiraoka, Y. (2017). Persistence diagrams with linear machine learning models.
arXiv preprint arXiv:1706.10082.

Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M.,
Prettenhofer, P., Weiss, R., Dubourg, V., et al. (2011). Scikit-learn: Machine learning in
python. Journal of Machine Learning Research, 12(Oct):2825–2830.

42
Penrose, M. D. and Yukich, J. E. (2001). Central limit theorems for some graphs in computational
geometry. Annals of Applied probability, pages 1005–1041.

Petrunin, A. (2007). Semiconcave functions in Alexandrov’s geometry. In Surveys in differential

geometry. Vol. XI, pages 137–201. Int. Press, Somerville, MA.

Phillips, J. M., Wang, B., and Zheng, Y. (2014). Geometric inference on kernel density estimates.
arXiv preprint 1307.7760.

Polonik, W. (1995). Measuring mass concentrations and estimating density contour clusters-an
excess mass approach. The Annals of Statistics, pages 855–881.

Poulenard, A., Skraba, P., and Ovsjanikov, M. (2018). Topological function optimization for
continuous shape matching. In Computer Graphics Forum, volume 37, pages 13–25. Wiley
Online Library.

Qaiser, T., Tsang, Y.-W., Taniyama, D., Sakamoto, N., Nakane, K., Epstein, D., and Rajpoot, N.
(2019). Fast and accurate tumor segmentation of histology images using persistent homology
and deep convolutional features. Medical image analysis, 55:1–14.

Ramamurthy, K. N., Varshney, K., and Mody, K. (2019). Topological data analysis of deci-
sion boundaries with application to model selection. In International Conference on Machine
Learning, pages 5351–5360. PMLR.

Reininghaus, J., Huber, S., Bauer, U., and Kwitt, R. (2015). A stable multi-scale kernel for
topological machine learning. In Proceedings of the IEEE Conference on Computer Vision and
Pattern Recognition, pages 4741–4748.

Rieck, B., Yates, T., Bock, C., Borgwardt, K., Wolf, G., Turk-Browne, N., and Krishnaswamy, S.
(2020). Uncovering the topology of time-varying fmri data using cubical persistence. Advances
in neural information processing systems, 33.

Rieck, B. A., Togninalli, M., Bock, C., Moor, M., Horn, M., Gumbsch, T., and Borgwardt, K.
(2019). Neural persistence: A complexity measure for deep neural networks using algebraic
topology. In International Conference on Learning Representations (ICLR 2019). OpenReview.

Robins, V. (1999). Towards computing homology from finite approximations. In Topology pro-
ceedings, volume 24, pages 503–532.

Robinson, A. and Turner, K. (2017). Hypothesis testing for topological data analysis. Journal
of Applied and Computational Topology, 1(2):241–261.

Roycraft, B., Krebs, J., and Polonik, W. (2020). Bootstrapping persistent betti numbers and
other stabilizing statistics. arXiv preprint arXiv:2005.01417.

Royer, M., Chazal, F., Levrard, C., Ike, Y., and Umeda, Y. (2021). Atol: Measure vectorisa-
tion for automatic topologically-oriented learning. In International Conference on Artificial
Intelligence and Statistics. PMLR.

Seversky, L. M., Davis, S., and Berger, M. (2016). On time-series topological data analysis:
new data and opportunities. In Proceedings of the IEEE Conference on Computer Vision and
Pattern Recognition Workshops, pages 59–67.

Singh, A., Scott, C., and Nowak, R. (2009). Adaptive Hausdorff estimation of density level sets.
Ann. Statist., 37(5B):2760–2782.

Singh, G., Mémoli, F., and Carlsson, G. E. (2007). Topological methods for the analysis of high
dimensional data sets and 3d object recognition. In SPBG, pages 91–100. Citeseer.

43
Skraba, P., Ovsjanikov, M., Chazal, F., and Guibas, L. (2010). Persistence-based segmentation
of deformable shapes. In Computer Vision and Pattern Recognition Workshops (CVPRW),
2010 IEEE Computer Society Conference on, pages 45–52.

Tsybakov, A. B. et al. (1997). On nonparametric estimation of density level sets. The Annals of
Statistics, 25(3):948–969.

Turner, K., Mileyko, Y., Mukherjee, S., and Harer, J. (2014a). Fréchet means for distributions
of persistence diagrams. Discrete & Computational Geometry, 52(1):44–70.

Turner, K., Mukherjee, S., and Boyer, D. M. (2014b). Persistent homology transform for modeling
shapes and surfaces. Information and Inference: A Journal of the IMA, 3(4):310–344.

Umeda, Y. (2017). Time series classification via topological data analysis. Transactions of the
Japanese Society for Artificial Intelligence, 32(3):D–G72_1.

Villani, C. (2003). Topics in Optimal Transportation. American Mathematical Society.

Walt, S. v. d., Colbert, S. C., and Varoquaux, G. (2011). The numpy array: a structure for
efficient numerical computation. Computing in Science & Engineering, 13(2):22–30.

Yao, Y., Sun, J., Huang, X., Bowman, G. R., Singh, G., Lesnick, M., Guibas, L. J., Pande, V. S.,
and Carlsson, G. (2009). Topological methods for exploring low-density states in biomolecular
folding pathways. The Journal of chemical physics, 130(14):144115.

Zieliński, B., Lipiński, M., Juda, M., Zeppelzauer, M., and Dłotko, P. (2010). Persistence bag-
of-words for topological data analysis. In Twenty-Eighth International Joint Conference on
Artificial Intelligence, IJCAI-19.

Zomorodian, A. and Carlsson, G. (2005). Computing persistent homology. Discrete Comput.

Geom., 33(2):249–274.

44