Data Exploration & Data Preprocessing

 Data Visualization

 Data Quality

 Major Tasks in Data Preprocessing

 Data Cleaning

 Data Integration

 Data Reduction

 Data Transformation

 Data Discretization & Concept Hierarchy generation

1
 Data Visualization
 Why data visualization?
 Gain insight into an information space by mapping data onto graphical
primitives
 Provide qualitative overview of large data sets
 Search for patterns, trends, structure, irregularities, relationships among
data
 Help to find interesting regions and suitable parameters for further
quantitative analysis
 Provide a visual proof of patterns derived through analytics / mining

2
 Example: Sea Surface Temperature
 The picture shows the Sea Surface Temperature
(SST) at different locations over the globe
 Tens of thousands of data points are

summarized in a single figure

Categorization of visualization methods:

• Pixel-oriented visualization techniques

• Geometric projection visualization
techniques
• Icon-based visualization techniques
• Hierarchical visualization techniques
• Visualizing complex data and relations

4
 Pixel-Oriented Visualization Techniques
 For a data set of m dimensions, sort the records as per a global order
 Create m windows on the screen, one for each dimension
 The m dimension values of a record are mapped to the corresponding
windows at the pixels at its global position based on the query
 The colors of the pixels reflect the corresponding values
 Figure high-lights the correlation between income and other attributes

(a) Income (b) Credit Limit (c) transaction volume (d) age
5
 Laying Out Pixels in Circle Segments
 A record is represented by a circle with colored segments of intensity
in accordance with the value of the attribute
 The pixels representing first chunk of records are located closest to the
center, followed by the latter records arranged outward in a continuum
in each sector corresponding to attributes.

Laying out pixels in circle segment

6
Dataset presented in Circle segments
Geometric Projection Visualization Techniques

 Visualization of geometric transformations and projections

of the multi-dimensional data
 Methods
 Scatterplot and scatterplot matrices
 Parallel coordinates

8
Scatter Plot Array of Iris Attributes
 Parallel Coordinates
 Presents k-D data along k equidistant axes which are parallel to one of
the screen axes and each corresponds to an attribute
 The axes are scaled to the [minimum, maximum]: range of the
corresponding attribute
 Every data item corresponds to a polygonal line which intersects each
of the axes at the point which corresponds to the value for the
attribute

• • •

Attr. 1 Attr. 2 Attr. 3 Attr. k

10
Parallel Coordinates of
a 3-D Labelled Data Set

11
Parallel Coordinates Plots for Iris Data
 Icon-Based Visualization Techniques

 Visualization of the data values as features of icons

 Typical visualization methods
 Chernoff Faces
 Stick Figures (5-piece)
 General techniques
 Shape coding: Use shape to represent certain
information encoding
 Color icons: Use color icons to encode more information
 Tile bars: Use small icons to represent the relevant
feature vectors in document retrieval

13
 Chernoff Faces
 A way to display variables on a two-dimensional surface, e.g., let x be
eyebrow slant, y be eye size, z be nose length, etc.
 The figure shows faces produced using 10 characteristics--head
eccentricity, eye size, eye spacing, eye eccentricity, pupil size,
eyebrow slant, nose size, mouth shape, mouth size, and mouth
opening): Each assigned one of 10 possible values, generated using
Mathematica (S. Dickson)

14
 Hierarchical Visualization Techniques

 So far the visualization techniques displayed

multiple dimensions simultaneously. But such
techniques have limited capacity to handle high
dimensional datasets.
 Hierarchical techniques partition the dimensions
into subsets and the resulting subspaces are
visualized in a hierarchical manner.
 Methods
 Worlds-within-Worlds
 Tree-Map

15
 Worlds-within-Worlds

 Assign the function(F) and two most important

parameters to innermost world
 Fix all other parameters at constant values - draw
other (1 or 2 or 3 dimensional worlds choosing
these as the axes)
 Software that uses this paradigm
 N–vision: Dynamic interaction through data glove
and stereo displays, including rotation, scaling
(inner) and translation (inner/outer)

16
 Worlds-within-worlds:Example
 In order to visualize the variation of feature F with reference to the
remaining five dimensions, X1,X2,..., X5, the values of the last
three dimensions will be fixed at say a, b, and c and visualize the
variation in F w.r.t X1 and X2 as 3-D plot, which is called as an
inner world located at origin (a,b,c) in the outer world defined by
(X3,X4,X5).
 Tree-Map
Screen-filling method which
uses a hierarchical
partitioning of the screen
into nested rectangular
regions depending on the
attribute values
The x- and y-dimension of
the screen are partitioned
alternately according to the
attribute values (classes)
Google news stories
organized into 7 categories,
in turn into sub-categories.
Ack.: http://www.cs.umd.edu/hcil/treemap-history/all102001.jpg 18
 Visualizing Complex Data and Relations
 Visualizing non-numerical data: text and social networks
 Tag cloud: visualizing user-generated tags in a web-site
 The importance of
tag is represented
by font size/color
 Besides text data,
there are also
methods to visualize
relationships, such as
visualizing social
networks

Newsmap: Google News Stories in 2005

Complex Relations among Entities-
Disease Influence graph

20
 Data Exploration & Preprocessing
 Data Visualization

 Data Quality

 Major Tasks in Data Preprocessing

 Data Cleaning

 Data Integration

 Data Reduction

 Data Transformation

 Data Discretization & Concept Hierarchy

21
 Data Quality: Why Preprocess the Data?

 Measures for data quality: A multidimensional view

 Accuracy: correct or wrong, accurate or not
 Completeness: not recorded, unavailable, …
 Consistency: some modified but some not, dangling, …
 Timeliness: timely update?
 Believability: how trustable the data are correct?
 Interpretability: how easily the data can be
understood?

22
 Data Exploration & Preprocessing
 Data Visualization

 Data Quality

 Major Tasks in Data Preprocessing

 Data Cleaning

 Data Integration

 Data Reduction

 Data Transformation

 Data Discretization & Concept Hierarchy generation

23
 Major Tasks in Data Preprocessing
 Data cleaning
 Fill in missing values, smooth noisy data, identify or remove
outliers, and resolve inconsistencies
 Data integration
 Integration of multiple databases, data cubes, or files
 Data reduction
 Dimensionality reduction
 Numerosity reduction
 Data compression
 Data transformation and data discretization
 Normalization
 Concept hierarchy generation

24
 Chapter 3: Data Preprocessing

 Data Visualization

 Data Quality

 Major Tasks in Data Preprocessing

 Data Cleaning

 Data Integration

 Data Reduction

 Data Transformation

 Data Discretization & Concept Hierarchy generation

25
 Data Cleaning
 Data in the Real World Is Dirty: Lots of potentially incorrect data,
e.g., instrument faulty, human or computer error, transmission error
 incomplete: lacking attribute values, lacking certain attributes of
interest, or containing only aggregate data
 e.g., Occupation=“ ” (missing data)
 noisy: containing noise, errors, or outliers
 e.g., Salary=“−10” (an error)
 inconsistent: containing discrepancies in codes or names, e.g.,
 Age=“42”, Birthday=“03/07/2010”
 Was rating “1, 2, 3”, now rating “A, B, C”
 discrepancy between duplicate records
 Intentional (e.g., disguised missing data)
 Jan. 1 as everyone’s birthday?
26
 Incomplete (Missing) Data

 Data is not always available

 E.g., many tuples have no recorded value for several
attributes, such as customer income in sales data
 Missing data may be due to
 equipment malfunction
 inconsistent with other recorded data and thus deleted
 data not entered due to misunderstanding
 certain data may not be considered important at the
time of entry
 Didn’t register history or changes of the data
 Missing data may need to be inferred
27
 How to Handle Missing Data?
 Ignore the tuple: usually done when class label is missing
(when doing classification)—not effective when the % of
missing values per attribute varies considerably
 Fill in the missing value manually: tedious + infeasible?
 Fill in it automatically with
 a global constant : e.g., “unknown”, a new class?!
 the attribute mean
 the attribute mean for all samples belonging to the
same class: smarter
 the most probable value: inference-based such as
Bayesian formula or decision tree
28
 Noisy Data
 Noise: random error or variance in a measured variable
 Incorrect attribute values may be due to
 faulty data collection instruments

 data entry problems

 data transmission problems

 technology limitation

 inconsistency in naming convention

29
 How to Handle Noisy Data?

 Binning
 first sort data and partition into (equal-frequency) bins

 then one can smooth by bin means, smooth by bin

median, smooth by bin boundaries, etc.

 Regression
 smooth by fitting the data into regression functions

 Noisy points / outlier detection

 Cluster formation to detect and remove outliers

 Combined computer and human inspection

 Auto-detect suspicious values and check by human

(e.g., deal with possible outliers)

30
 Binning Methods for Data Smoothing
 Sorted data for price (in dollars): 4, 8, 9, 15, 21, 21, 24, 25, 26,
28, 29, 34
* Partition into equal-frequency (equi-depth) bins:
- Bin 1: 4, 8, 9, 15
- Bin 2: 21, 21, 24, 25
- Bin 3: 26, 28, 29, 34
* Smoothing by bin means:
- Bin 1: 9, 9, 9, 9
- Bin 2: 23, 23, 23, 23
- Bin 3: 29, 29, 29, 29
* Smoothing by bin boundaries:
- Bin 1: 4, 4, 4, 15
- Bin 2: 21, 21, 25, 25
- Bin 3: 26, 26, 26, 34
31
 y
Regression Analysis
Y1

 Regression analysis: A collective name for

techniques for the modeling and analysis Y1’
y=x+1
of numerical data consisting of values of a
dependent variable (also called
response variable or measurement) and X1 x
of one or more independent variables (aka.
explanatory variables or predictors)
 Used for prediction
 The parameters are estimated so as to give (including forecasting of
a "best fit" of the data time-series data), inference,
hypothesis testing, and
 Most commonly the best fit is evaluated by
modeling of causal
using the least squares method, but
relationships
other criteria have also been used

32
 Chapter 3: Data Preprocessing

 Data Preprocessing: An Overview

 Data Quality

 Major Tasks in Data Preprocessing

 Data Cleaning

 Data Integration

 Data Reduction

 Data Transformation and Data Discretization

 Summary
33
 Data Integration
 Data integration:
 Combines data from multiple sources into a coherent store
 Schema integration: e.g., A.cust-id  B.cust-#
 Integrate metadata from different sources
 Entity identification problem:
 Identify same real world entities from multiple data sources,
 e.g., S.V.N.Rao =S.V.Narayan Rao = S.Venkat Narayana Rao

 Detecting and resolving data value conflicts

 For the same real world entity, attribute values from different
sources are different e.g., distance between cities
 Possible reasons: different representations, precision levels,
metrics, scales, e.g., American units vs. British units
34
 Handling Redundancy in Data Integration

 Redundant data occur often when integration of multiple

databases
 Object identification: The same attribute or object
may have different names in different databases
 Derivable data: One attribute may be a “derived”
attribute of another attribute, e.g., annual revenue
 Redundant attributes can be detected by correlation
analysis and covariance analysis
 Proper integration methods aim to reduce/avoid
redundancies and inconsistencies leading to improved
mining speed and quality
35
 Correlation Analysis (Nominal Data)

(Observed  Expected ) 2
2  
Expected

36
 Chi-Square Calculation: An Example

Play chess Not play chess Sum (row)

Like science fiction 250(90) 200(360) 450

Not like science fiction 50(210) 1000(840) 1050

Sum(col.) 300 1200 1500

 Χ2 (chi-square) calculation (numbers in parenthesis are

expected counts calculated by dividing the product of the
marginal counts of the two categories by the total count)
(250  90) 2 (50  210) 2 (200  360) 2 (1000  840) 2
 
2
    507.93
90 210 360 840

 It shows that like_science_fiction and play_chess are

correlated in the group
37
 Correlation Analysis (Numeric Data)

 Correlation coefficient (also called Pearson’s product

moment coefficient)

i1 (ai  A)(bi  B) 

n n
(ai bi )  n AB
rA, B   i 1
n A B n A B

where n is the number of tuples, A and B are the respective

means of A and B, σA and σB are the respective standard deviation
of A and B, and Σ(aibi) is the sum of the product of A and B
variables for each tuple.
 If rA,B > 0, A and B are positively correlated (A’s values
increase as B’s). The higher, the stronger correlation.
 rA,B = 0: independent; rAB < 0: negatively correlated
38
Visually Evaluating Correlation

Scatter plots
showing the
correlation co-
efficient(rAB) from –
1 to 1.

39
 Covariance (Numeric Data)
 Covariance is similar to correlation

Correlation coefficient:

where n is the number of tuples, A and B are the respective mean or

expected values of A and B, σA and σB are the respective standard
deviation of A and B.
 Positive covariance: If CovA,B > 0, then A and B both tend to be larger
than their expected values.
 Negative covariance: If CovA,B < 0 then if A is larger than its expected
value, B is likely to be smaller than its expected value.
 Independence: CovA,B = 0 but the converse is not true:
 Some pairs of random variables may have a covariance of 0 but are not
independent. Only under some additional assumptions (e.g., the data follow
multivariate normal distributions) does a covariance of 0 imply independence40
 Co-Variance: An Example

 It can be simplified in computation as

 Suppose two stocks A and B have the following values in one week:
(2, 5), (3, 8), (5, 10), (4, 11), (6, 14).

 Question: Will the prices rise or fall together? If so, the stocks are
affected by the same industry trend.

 A= (2 + 3 + 5 + 4 + 6)/ 5 = 20/5 = 4

 B = (5 + 8 + 10 + 11 + 14) /5 = 48/5 = 9.6

 Cov(A,B) = (2×5+3×8+5×10+4×11+6×14)/5 − 4 × 9.6 = 4

 Thus, A and B rise together since Cov(A, B) > 0.

 Chapter 3: Data Preprocessing

 Data Preprocessing: An Overview

 Data Quality

 Major Tasks in Data Preprocessing

 Data Cleaning

 Data Integration

 Data Reduction

 Data Transformation and Data Discretization

 Summary
42
 Data Reduction Strategies
 Data reduction: Obtain a reduced representation of the data set that
is much smaller in volume but yet produces the same (or almost the
same) analytical results
 Why data reduction? — A database/data warehouse may store
terabytes of data. Complex data analysis may take a very long time to
run on the complete data set.
 Data reduction strategies
 Dimensionality reduction, e.g., remove unimportant attributes

 Wavelet transforms

 Principal Components Analysis (PCA)

 Feature subset selection, feature creation

 Numerosity reduction (some simply call it: Data Reduction)

 Regression and Log-Linear Models

 Histograms, clustering, sampling

 Data cube aggregation

 Data compression

43
Data Reduction 1: Dimensionality Reduction
 Curse of dimensionality
 When dimensionality increases, data becomes increasingly sparse
 Density and distance between points, which is critical to clustering and
outlier analysis, becomes less meaningful
 The possible combinations of subspaces will grow exponentially
 Dimensionality reduction
 Avoid the curse of dimensionality
 Help eliminate irrelevant features and reduce noise
 Reduce time and space required for data mining
 Allow easier visualization
 Dimensionality reduction techniques
 Wavelet transforms
 Principal Component Analysis
 Supervised and nonlinear techniques (e.g., feature selection)

44
 Mapping Data to a New Space
 Fourier transform
 Wavelet transform

Two Sine Waves Two Sine Waves + Noise Frequency

45
 Wavelet Transformation
 Discrete wavelet transform (DWT) is a linear signal
processing technique that transforms a given data vector X,
into a numerically different vector, X’ of wavelet coefficients
 Useful for data reduction by Compressed approximations:
store only a small fraction of the strongest of the wavelet
coefficients by thresholding
 Provides multi-resolution analysis and denoising & cleaning
 Similar to discrete Fourier transform (DFT), but better lossy
compression, localized in space

46
 What Is Wavelet Transform?

Haar2 Daubechie4
 Several Families of DWTs exist and two of them are shown in the above
picture
 General procedure is the Hierarchical Pyramid Algorithm:
 Length, L, must be an integer power of 2 (padding with 0’s, when
necessary)
 Each transform has 2 functions:
 f1: smoothing for aggregation
 f2: difference for details
 Applies to pairs of successive data points, resulting in two set of data of
length L/2 corresponding to low and high frequency content
 Applies two functions recursively, until the length of each list reaches 2
47
 Why Wavelet Transform?

 Data are transformed to preserve relative distance

between objects at different levels of resolution
 Allow natural clusters to become more distinguishable
 Effective removal of outliers
 Used for data cleaning to remove noise

 insensitive to input order

 More effective than JPEG compression
 Efficient with time complexity O(N)
 Real world applications includes compression of finger
print images, computer vision, time-series data analysis

48
 Principal Component Analysis (PCA)
 Find a projection that captures the largest amount of variation in data
 The original data are projected onto a much smaller space, resulting
in dimensionality reduction. We find the eigenvectors of the
covariance matrix, and these eigenvectors define the new space

x2

x1
49
 Principal Component Analysis (Steps)
 Given N data vectors from n-dimensions, find k ≤ n orthogonal vectors
(principal components) that can be best used to represent data
 Normalize input data: Each attribute falls within the same range
 Compute k orthonormal (unit) vectors, i.e., principal components
 The principal components are sorted in the order of decreasing
“significance” or strength
 Each input data (vector) is represented in a low dimensional space
defined by the k principal component vectors
 Since the components are sorted, the size of the data can be reduced
by eliminating the weak components, i.e., those with low variance
(i.e., using the strongest principal components, it is possible to
reconstruct a good approximation of the original data)
 Works for numeric data only
50
 Attribute Subset Selection
 Another way to reduce dimensionality of data
 Redundant attributes
 Duplicate the information contained in one or more
other attributes
 E.g., price of a product and the sales tax paid
 Irrelevant attributes
 Contain no information that is useful for the data
mining task at hand
 E.g., students' ID is often irrelevant to the task of
predicting students' CGPA

51
 Heuristic Search in Attribute Selection

 There are 2d possible attribute combinations of d attributes

 Typical heuristic attribute selection methods:
 under the attribute independence assumption, impact of attributes

is assessed by significance tests and best attributes chosen

1. Step-wise forward selection:
 The best single-attribute is picked first
 Then the next best attribute, condition to the first, ...
2. Step-wise backward elimination:
 Repeatedly eliminate the worst attribute

 Combined attribute selection and elimination:

 At each step select the best and eliminate the worst attribute

refining the selected set of k-attributes iteratively

 Decision Tree Induction:

 All attributes that appear in the tree are selected and the rest

eliminated
52
Attribute Creation (Feature Generation)
 Create new attributes (features) that can capture the
important information in a data set more effectively than the
original ones
 Attribute construction

 New attributes are created by Combining features

 Examples:

 generate new attributes like area, density, etc.

 score for multi-criteria based ranking

53
Data Reduction 2: Numerosity Reduction
 Reduce data volume by choosing alternative, smaller
forms of data representation
 Parametric methods (e.g., regression)
 Assume the data fits some model, estimate model

parameters, store only the parameters, and discard

the data (except possible outliers)
 Log-linear models—obtain value at a point in m-D

space as a combination of multiple lower-dimensional

spaces
 Non-parametric methods
 Do not assume data distribution models

 Major families: histograms, clustering, sampling, …

54
Regress Analysis and Log-Linear Models
 Linear regression: Y = w X + b
 Two regression coefficients, w and b, specify the line and are to be
estimated by using the data at hand
 Using the least squares criterion to the known values of Y1, Y2, …,
X1, X2, ….
 Multiple regression: Y = b0 + b1 X1 + b2 X2
 Many nonlinear functions can be transformed into the above
 Log-linear models:
 Approximate discrete multidimensional probability distributions
 Estimate the probability of each point (tuple) in a multi-dimensional
space for a combination of discretized attributes, based on a
smaller subsets of dimensional combinations
 Useful for data smoothing and also for dimensionality reduction
55
 Histogram Analysis
 Divide data into buckets and
store average (sum) for each
bucket
 Partitioning rules:
 Equal-width: equal bucket
range
 Equal-frequency (or equal-
depth)

56
 Clustering
 Partition data set into clusters based on similarity, and
store cluster representatives (e.g., centroid and
diameter) only
 Can be very effective if data is clusterable but not
effective otherwise (if the data is “smeared”)
 Can have hierarchical clustering and be stored in multi-
dimensional index tree structures

57
 Sampling

 Sampling: obtaining a small sample s to represent the whole data set

N
 Allow a mining algorithm to run in complexity that is potentially sub-
linear to the size of the data
 Key principle: Choose a representative subset of the data
 Simple random sampling may have very poor performance in the
presence of skew
 Develop adaptive sampling methods, e.g., stratified sampling:
 Sampling may not reduce database I/Os (page at a time)
 Solution is cluster sampling wherein each cluster contains a page/
sector of records and k clusters are selected at random from n-
clusters (k << n) constituting a database.

58
 Types of Sampling
 Simple random sampling
 There is an equal probability of selecting any particular
item from a given population
 Sampling without replacement

 Once an item is selected, it is removed from the population.

 Results in a sample of distinct items
 Sampling with replacement
 A selected object is not removed from the population.
 Results in a sample containing possibly repeated items
 Stratified sampling:
 Partition the data set, and draw samples from each
partition proportionally.
 sample has approximately the same percentage of items
representing different groups in the data
 Used in conjunction with skewed data
59
Sampling: With or without Replacement

Raw Data
60
Sampling: Stratified Sampling

Raw Data Stratified Sample

61
 Data Cube Aggregation

 The lowest level of a data cube (base cuboid)

 The aggregated data for an individual entity of interest
 E.g., a customer in a phone calling data warehouse
 Multiple levels of aggregation in data cubes
 Further reduce the size of data to deal with
 Refer appropriate levels
 Use the smallest representation which is enough to
solve the task
 Queries regarding aggregated information should be
answered using data cube, when possible
62
 Data Compression

Original Data Compressed

Data
lossless

Original Data
Approximated

63
 Data Compression

 Transformations are applied to obtain a reduced

or compressed representation of the original
data.
 If the original data can be reconstructed from the
compressed data, it is lossless compression.
 If there would be information loss in the
reconstructed data, it is called lossy compression.
 Dimensionality reduction techniques (Wavelets)
and some of the numerosity reduction techniques
(clustering) discussed are considered as different
forms of data compression.
 Chapter 3: Data Preprocessing

 Data Preprocessing: An Overview

 Data Quality

 Major Tasks in Data Preprocessing

 Data Cleaning

 Data Integration

 Data Reduction

 Data Transformation and Data Discretization

 Summary
65
 Data Transformation
 A function that maps the entire set of values of a given attribute to a new
set of replacement values such that each old value can be identified with
one of the new values
 Methods
 Smoothing: Remove noise from data
 Attribute/feature construction
 New attributes constructed from the given ones
 Aggregation: Summarization, data cube construction
 Normalization: Scaled to fall within a smaller, specified range
 min-max normalization
 z-score normalization
 normalization by decimal scaling
 Discretization / Concept hierarchy generation
66
 Normalization
 Min-max normalization: to [new_minA, new_maxA]
v  minA
v'  (new _ maxA  new _ minA)  new _ minA
maxA  minA
 Ex. Let income range $12,000 to $98,000 normalized to [0.0,
73,600  12,000
1.0]. Then $73,600 is mapped to 98,000 12,000 (1.0  0)  0  0.716
 Z-score normalization (μ: mean, σ: standard deviation):
v  A
v' 
 A

73,600  54,000
 Ex. Let μ = 54,000, σ = 16,000. Then  1.225
16,000
 Normalization by decimal scaling
v
v' j Where j is the smallest integer such that Max(|ν’|) < 1
10
67
 Data Discretization
 Discretization: Divide the range of a continuous attribute into
intervals and the intervals can be labelled
 Interval labels are used to replace actual data values so that
numeric features are transformed into nominal features for further
analysis, e.g., classification
 Typical methods: All the methods can be applied recursively
 Binning: Top-down split, unsupervised
 Histogram analysis: Top-down split, unsupervised
 Clustering analysis: Unsupervised, top-down split or bottom-up
merge)
 Decision-tree analysis (supervised, top-down split)
 Correlation (e.g., 2) analysis (supervised, bottom-up merge)
(adjacent intervals having similar class distribution are merged)
68
 Simple Discretization: Binning

 Equal-width (distance) partitioning

 Divides the range into N intervals of equal size: uniform grid
 if A and B are the lowest and highest values of the attribute, the
width of intervals will be: W = (B –A)/N.
 The most straightforward, but outliers may dominate presentation
 Skewed data is not handled well

 Equal-depth (frequency) partitioning

 Divides the range into N intervals, each containing approximately
same number of samples
 Good data scaling
 Managing categorical attributes can be tricky
69
Discretization Without Using Class Labels
(Binning vs. Clustering)

Data Equal width (binning)

Equal interval width (binning)

Equal frequency (binning) K-means clustering leads to better results

70
 Discretization by Classification &
Correlation Analysis
 Classification (Decision tree analysis)
 Supervised: Given class labels, e.g., cancerous vs. benign
 Using entropy to determine split points (discretization point)
 Top-down, recursive split
 Correlation analysis (Chi-merge: χ2-based discretization)
 Supervised: use class information
 Bottom-up merge: find the best neighboring intervals (those
having similar distributions of classes, i.e., low χ2 values) to merge
 Merge performed recursively, until a predefined stopping condition

71
 Concept Hierarchy Generation
 Concept hierarchy organizes features hierarchically and is usually
associated with each dimension in a data warehouse
 Concept hierarchies facilitate drilling down and rolling up to view /
analyze data (maintained in a data warehouse) at multiple granularity
 Concept hierarchy formation:
 Recursively reduce the data by collecting and replacing low level
values / concepts (such as numeric values for age) by higher level
concepts (such as youth, adult, or senior)
 Concept hierarchies can be explicitly specified by domain experts
and/or data warehouse designers OR
 Concept hierarchy can be automatically formed for both numeric and
nominal data.
 For numeric data, discretization methods are used.

72
 Concept Hierarchy Generation
for Nominal Data
 Specification of a partial/total ordering of attributes explicitly at
the schema level by users or experts
 street < city < state < country
 Specification of a hierarchy for a set of values by explicit data
grouping
 {Vizag, Vijayawada, Tirupati} < AP
 Specification of a set of attributes only that constitute the
concept hierarchy but not their ordering
 E.g., for a set of attributes: {street, city, state, country}
 Solution:Automatic generation of hierarchies (or attribute
levels) by analyzing a dataset based on the number of
distinct values
73
 Automatic Concept Hierarchy Generation
 Some hierarchies can be automatically generated based on
the analysis of the number of distinct values per attribute in
the data set
 The attributes with more number of distinct values are
placed at the lower levels of the hierarchy
 Exceptions, e.g., weekday, month, quarter, year

country 15 distinct values

province_or_ state 365 distinct values

city 3567 distinct values

street 674,339 distinct values

74
 Chapter 3: Data Preprocessing

 Data Preprocessing: An Overview

 Data Quality

 Major Tasks in Data Preprocessing

 Data Cleaning

 Data Integration

 Data Reduction

 Data Transformation and Data Discretization

 Summary
75
 Summary
 Data quality: accuracy, completeness, consistency, timeliness,
believability, interpretability
 Data cleaning: e.g. missing/noisy values, outliers
 Data integration from multiple sources:
 Entity identification problem

 Remove redundancies

 Detect inconsistencies

 Data reduction
 Dimensionality reduction

 Numerosity reduction

 Data compression

 Data transformation and data discretization

 Normalization

 Concept hierarchy generation

76