Note to other teachers and users of

these slides. Andrew would be delighted

if you found this source material useful in
giving your own lectures. Feel free to use
these slides verbatim, or to modify them
to fit your own needs. PowerPoint
originals are available. If you make use
Introduction to Support
of a significant portion of these slides in
your own lecture, please include this
message, or the following link to the
Vector Machines
source repository of Andrew’s tutorials:
http://www.cs.cmu.edu/~awm/tutorials .
Comments and corrections gratefully
received.

Thanks:
Andrew Moore
CMU
And
Martin Law
Michigan State University
 History of SVM
 SVM is related to statistical learning theory [3]
 SVM was first introduced in 1992 [1]

 SVM becomes popular because of its success in

handwritten digit recognition
 1.1% test error rate for SVM. This is the same as the error
rates of a carefully constructed neural network, LeNet 4.
 See Section 5.11 in [2] or the discussion in [3] for details
 SVM is now regarded as an important example of “kernel
methods”, one of the key area in machine learning
 Note: the meaning of “kernel” is different from the “kernel”
function for Parzen windows
[1] B.E. Boser et al. A Training Algorithm for Optimal Margin Classifiers. Proceedings of the Fifth Annual Workshop on
Computational Learning Theory 5 144-152, Pittsburgh, 1992.
[2] L. Bottou et al. Comparison of classifier methods: a case study in handwritten digit recognition. Proceedings of the 12th
IAPR International Conference on Pattern Recognition, vol. 2, pp. 77-82.
[3] V. Vapnik. The Nature of Statistical Learning Theory. 2nd edition, Springer, 1999.

2021/3/3 2
Linear Classifiers Estimation:
x f yest
f(x,w,b) = sign(w. x - b)
denotes +1
denotes -1
w: weight vector
x: data vector

How would you

classify this data?

2021/3/3 3
 a
Linear Classifiers
x f yest
f(x,w,b) = sign(w. x - b)
denotes +1
denotes -1

How would you

classify this data?

2021/3/3 4
 a
Linear Classifiers
x f yest
f(x,w,b) = sign(w. x - b)
denotes +1
denotes -1

How would you

classify this data?

2021/3/3 5
 a
Linear Classifiers
x f yest
f(x,w,b) = sign(w. x - b)
denotes +1
denotes -1

How would you

classify this data?

2021/3/3 6
 a
Linear Classifiers
x f yest
f(x,w,b) = sign(w. x - b)
denotes +1
denotes -1

Any of these
would be fine..

..but which is
best?

2021/3/3 7
 a
Classifier Margin
x f yest
f(x,w,b) = sign(w. x - b)
denotes +1
denotes -1 Define the margin
of a linear
classifier as the
width that the
boundary could be
increased by
before hitting a
datapoint.

2021/3/3 8
 a
Maximum Margin
x f yest
f(x,w,b) = sign(w. x - b)
denotes +1
denotes -1 The maximum
margin linear
classifier is the
linear classifier
with the, um,
maximum margin.
This is the
simplest kind of
SVM (Called an
LSVM)
Linear SVM
2021/3/3 9
 a
Maximum Margin
x f yest
f(x,w,b) = sign(w. x + b)
denotes +1
denotes -1 The maximum
margin linear
classifier is the
linear classifier
Support Vectors with the, um,
are those
datapoints that maximum margin.
the margin This is the
pushes up
against simplest kind of
SVM (Called an
LSVM)
Linear SVM
2021/3/3 10
 Why Maximum Margin?

f(x,w,b) = sign(w. x - b)
denotes +1
denotes -1 The maximum
margin linear
classifier is the
linear classifier
Support Vectors with the, um,
are those
datapoints that maximum margin.
the margin This is the
pushes up
against simplest kind of
SVM (Called an
LSVM)

2021/3/3 11
How to calculate the distance from a point
to a line?
denotes +1
denotes -1 x
wx +b = 0

X – Vector
W
W – Normal Vector
b – Scale Value

 http://mathworld.wolfram.com/Point-LineDistance2-
Dimensional.html
 In our case, w1*x1+w2*x2+b=0,

 thus, w=(w1,w2), x=(x1,x2)

2021/3/3 12
Estimate the Margin
denotes +1
denotes -1 x
wx +b = 0

X – Vector
W
W – Normal Vector
b – Scale Value

 What is the distance expression for a point x to a line

wx+b= 0?
xw b xw b
d ( x)  

2 d 2
w w
i 1 i
2

2021/3/3 13
Large-margin Decision Boundary
 The decision boundary should be as far away from the
data of both classes as possible
 We should maximize the margin, m

 Distance between the origin and the line wtx=-b is b/||w||

Class 2

Class 1
m

2021/3/3 14
Finding the Decision Boundary
 Let {x1, ..., xn} be our data set and let yi  {1,-1} be
the class label of xi
 The decision boundary should classify all points correctly

 To see this: when y=-1, we wish (wx+b)<1, when y=1,

we wish (wx+b)>1. For support vectors, we wish

y(wx+b)=1.
 The decision boundary can be found by solving the
following constrained optimization problem

2021/3/3 15
Next step… Optional
 Converting SVM to a form we can solve
 Dual form
 Allowing a few errors
 Soft margin
 Allowing nonlinear boundary
 Kernel functions

2021/3/3 16
 The Dual Problem (we ignore the derivation)
 The new objective function is in terms of ai only
 It is known as the dual problem: if we know w, we

know all ai; if we know all ai, we know w

 The original problem is known as the primal problem

 The objective function of the dual problem needs to be

maximized!
 The dual problem is therefore:

Properties of ai when we introduce The result when we differentiate the

the Lagrange multipliers original Lagrangian w.r.t. b
2021/3/3 17
The Dual Problem

 This is a quadratic programming (QP) problem

 A global maximum of ai can always be found

 w can be recovered by

2021/3/3 18
Characteristics of the Solution
 Many of the ai are zero (see next page for example)
w is a linear combination of a small number of data points
 This “sparse” representation can be viewed as data

compression as in the construction of knn classifier

 xi with non-zero ai are called support vectors (SV)

 The decision boundary is determined only by the SV

 Let tj (j=1, ..., s) be the indices of the s support vectors.

We can write
 For testing with a new data z
 Compute and
classify z as class 1 if the sum is positive, and class 2
otherwise
 Note: w need not be formed explicitly
2021/3/3 19
A Geometrical Interpretation

Class 2

a8=0.6 a10=0

a7=0
a5=0 a2=0

a1=0.8
a4=0
a6=1.4
a9=0
a3=0
Class 1

2021/3/3 20
Allowing errors in our solutions
We allow “error” xi in classification; it is based on the
output of the discriminant function wTx+b
 xi approximates the number of misclassified samples

Class 2

Class 1

2021/3/3 21
Soft Margin Hyperplane
 If we minimize ixi, xi can be computed by

 xi are “slack variables” in optimization

 Note that xi=0 if there is no error for xi

 xi is an upper bound of the number of errors

 We want to minimize
C : tradeoff parameter between error and margin


 The optimization problem becomes

2021/3/3 22
Extension to Non-linear Decision Boundary
So far, we have only considered large-
margin classifier with a linear decision
boundary
How to generalize it to become nonlinear?

Key idea: transform xi to a higher

dimensional space to “make life easier”
 Input space: the space the point xi are
located
 Feature space: the space of f(xi) after

transformation
2021/3/3 23
 Transforming the Data (c.f. DHS Ch. 5)
f( )
f( ) f( )
f( ) f( ) f( )
f(.) f( )
f( ) f( )
f( ) f( )
f( ) f( )
f( ) f( ) f( )
f( )
f( )

Input space Feature space

Note: feature space is of higher dimension
than the input space in practice

 Computation in the feature space can be costly because it is

high dimensional
 The feature space is typically infinite-dimensional!
 The kernel trick comes to rescue

2021/3/3 24
The Kernel Trick
 Recall the SVM optimization problem

 The data points only appear as inner product

 As long as we can calculate the inner product in the

feature space, we do not need the mapping explicitly

 Many common geometric operations (angles, distances)
can be expressed by inner products
 Define the kernel function K by

2021/3/3 25
An Example for f(.) and K(.,.)
 Suppose f(.) is given as follows

 An inner product in the feature space is

 So, if we define the kernel function as follows, there is

no need to carry out f(.) explicitly

 This use of kernel function to avoid carrying out f(.)

explicitly is known as the kernel trick

2021/3/3 26
More on Kernel Functions
 Not all similarity measures can be used as kernel
function, however
 The kernel function needs to satisfy the Mercer function,
i.e., the function is “positive-definite”
 This implies that
 the n by n kernel matrix,
 in which the (i,j)-th entry is the K(xi, xj), is always positive
definite
 This also means that optimization problem can be solved
in polynomial time!

2021/3/3 27
Examples of Kernel Functions

 Polynomial kernel with degree d

 Radial basis function kernel with width s

 Closely related to radial basis function neural networks

 The feature space is infinite-dimensional

 Sigmoid with parameter k and q

 It does not satisfy the Mercer condition on all k and q

2021/3/3 28
Non-linear SVMs: Feature spaces

 General idea: the original input space can always be mapped to

some higher-dimensional feature space where the training set is
separable:

Φ: x → φ(x)

2021/3/3 29
Example
 Suppose we have 5 one-dimensional data points
 x1=1, x2=2, x3=4, x4=5, x5=6, with 1, 2, 6 as class 1 and 4,
5 as class 2  y1=1, y2=1, y3=-1, y4=-1, y5=1
 We use the polynomial kernel of degree 2
 K(x,y) = (xy+1)2
 C is set to 100

 We first find ai (i=1, …, 5) by

2021/3/3 30
Example
 By using a QP solver, we get
 a1=0, a2=2.5, a3=0, a4=7.333, a5=4.833
 Note that the constraints are indeed satisfied

 The support vectors are {x2=2, x4=5, x5=6}

 The discriminant function is

 b is recovered by solving f(2)=1 or by f(5)=-1 or by f(6)=1,

as x2 and x5 lie on the line and x4
lies on the line
 All three give b=9

2021/3/3 31
Example

Value of discriminant function

class 1 class 2 class 1

1 2 4 5 6

2021/3/3 32
Degree of Polynomial Features

X^1 X^2 X^3

X^4 X^5 X^6

2021/3/3 33
Choosing the Kernel Function
 Probably the most tricky part of using SVM.

2021/3/3 34
Software
 A list of SVM implementation can be found at
http://www.kernel-machines.org/software.html
 Some implementation (such as LIBSVM) can handle
multi-class classification
 SVMLight is among one of the earliest implementation of

SVM
 Several Matlab toolboxes for SVM are also available

2021/3/3 35
Summary: Steps for Classification
 Prepare the pattern matrix
 Select the kernel function to use

 Select the parameter of the kernel function and the

value of C
 You can use the values suggested by the SVM software, or
you can set apart a validation set to determine the values
of the parameter
 Execute the training algorithm and obtain the ai
 Unseen data can be classified using the ai and the

support vectors

2021/3/3 36
Conclusion
 SVM is a useful alternative to neural networks
 Two key concepts of SVM: maximize the margin and the

kernel trick
 Many SVM implementations are available on the web for

you to try on your data set!

2021/3/3 37
Resources
 http://www.kernel-machines.org/
 http://www.support-vector.net/

 http://www.support-vector.net/icml-tutorial.pdf

 http://www.kernel-machines.org/papers/tutorial-

nips.ps.gz
 http://www.clopinet.com/isabelle/Projects/SVM/applist.h
tml

2021/3/3 38
 Appendix: Distance from a point to a line

 Equation for the line: let u be a variable, then any point

on the line can be described as:
 P = P1 + u (P2 - P1)
 Let the intersect point be u, P2
 Then, u can be determined by:

 The two vectors (P2-P1) is orthogonal to P3-u:

P
 That is,

 (P3-P) dot (P2-P1) =0

 P=P1+u(P2-P1)
P3
 P1=(x1,y1),P2=(x2,y2),P3=(x3,y3) P1

2021/3/3 39
Distance and margin

 x = x1 + u (x2 - x1)
y = y1 + u (y2 - y1)

 The distance therefore between the point P3 and the

line is the distance between P=(x,y) above and P3
 Thus,

 d= |(P3-P)|=

2021/3/3 40