(10 intermediate revisions by 5 users not shown)
Line 1: Line 1:
[http://balthier.ecn.purdue.edu/index.php/ECE662 ECE662 Main Page]
+
=Lecture 9, [[ECE662]]: Decision Theory=
  
[http://balthier.ecn.purdue.edu/index.php/ECE662#Class_Lecture_Notes Class Lecture Notes]
+
Lecture notes for [[ECE662:BoutinSpring08_Old_Kiwi|ECE662 Spring 2008]], Prof. [[user:mboutin|Boutin]].
 +
 
 +
Other lectures: [[Lecture 1 - Introduction_Old Kiwi|1]],
 +
[[Lecture 2 - Decision Hypersurfaces_Old Kiwi|2]],
 +
[[Lecture 3 - Bayes classification_Old Kiwi|3]],
 +
[[Lecture 4 - Bayes Classification_Old Kiwi|4]],
 +
[[Lecture 5 - Discriminant Functions_Old Kiwi|5]],
 +
[[Lecture 6 - Discriminant Functions_Old Kiwi|6]],
 +
[[Lecture 7 - MLE and BPE_Old Kiwi|7]],
 +
[[Lecture 8 - MLE, BPE and Linear Discriminant Functions_Old Kiwi|8]],
 +
[[Lecture 9 - Linear Discriminant Functions_Old Kiwi|9]],
 +
[[Lecture 10 - Batch Perceptron and Fisher Linear Discriminant_Old Kiwi|10]],
 +
[[Lecture 11 - Fischer's Linear Discriminant again_Old Kiwi|11]],
 +
[[Lecture 12 - Support Vector Machine and Quadratic Optimization Problem_Old Kiwi|12]],
 +
[[Lecture 13 - Kernel function for SVMs and ANNs introduction_Old Kiwi|13]], 
 +
[[Lecture 14 - ANNs, Non-parametric Density Estimation (Parzen Window)_Old Kiwi|14]],
 +
[[Lecture 15 - Parzen Window Method_Old Kiwi|15]],
 +
[[Lecture 16 - Parzen Window Method and K-nearest Neighbor Density Estimate_Old Kiwi|16]],
 +
[[Lecture 17 - Nearest Neighbors Clarification Rule and Metrics_Old Kiwi|17]],
 +
[[Lecture 18 - Nearest Neighbors Clarification Rule and Metrics(Continued)_Old Kiwi|18]],
 +
[[Lecture 19 - Nearest Neighbor Error Rates_Old Kiwi|19]],
 +
[[Lecture 20 - Density Estimation using Series Expansion and Decision Trees_Old Kiwi|20]],
 +
[[Lecture 21 - Decision Trees(Continued)_Old Kiwi|21]],
 +
[[Lecture 22 - Decision Trees and Clustering_Old Kiwi|22]],
 +
[[Lecture 23 - Spanning Trees_Old Kiwi|23]],
 +
[[Lecture 24 - Clustering and Hierarchical Clustering_Old Kiwi|24]],
 +
[[Lecture 25 - Clustering Algorithms_Old Kiwi|25]],
 +
[[Lecture 26 - Statistical Clustering Methods_Old Kiwi|26]],
 +
[[Lecture 27 - Clustering by finding valleys of densities_Old Kiwi|27]],
 +
[[Lecture 28 - Final lecture_Old Kiwi|28]],
 +
----
 +
----
  
 
'''Parametric Methods'''
 
'''Parametric Methods'''
  
 
Two applications:
 
Two applications:
* Parametric Density Estimation: Using sample data, we estimate probabilities <math>P(\omega_i)</math>, <math>p(\vec{X}|\omega_i) \forall i</math> etc using estimation methods like [MLE] and [BPE]. Then from these estimated probabilities (and not true probabilities, which are unknown; only their parametric form was known) we can use Bayes classification rule to build a classifier.
+
* Parametric Density Estimation: Using sample data, we estimate probabilities <math>P(\omega_i)</math>, <math>p(\vec{X}|\omega_i) \forall i</math> etc using estimation methods like [[MLE_Old Kiwi]] and [[BPE_Old Kiwi]]. Then from these estimated probabilities (and not true probabilities, which are unknown; only their parametric form was known) we can use Bayes classification rule to build a classifier.
  
 
* [[Parametric Classifiers_Old Kiwi]]: We find parametric decision boundaries to approximate true decision boundaries between classes. This is very different approach from approximating the probabilities with their estimates, as in previous method.
 
* [[Parametric Classifiers_Old Kiwi]]: We find parametric decision boundaries to approximate true decision boundaries between classes. This is very different approach from approximating the probabilities with their estimates, as in previous method.
Line 20: Line 51:
 
E.g. 'g' can be a degree-2 polynomial, or it can be a degree-1 polynomial (resulting in a linear classifier).
 
E.g. 'g' can be a degree-2 polynomial, or it can be a degree-1 polynomial (resulting in a linear classifier).
  
If 'g' is linear, it can be written as <math>g(\vec{X})=\vec{V}.\vec{X}+V_0=(\vec{V},V_0).(\vec{X},1)</math> or <math>g(1,\vec{X})=\vec{c}.(1,\vec{X})</math>.
+
If 'g' is linear, it can be written as <math>g(\vec{X})=\vec{V}\cdot \vec{X}+V_0=(\vec{V},V_0)\cdot (\vec{X},1)</math> or <math>g(1,\vec{X})=\vec{c}\cdot(1,\vec{X})</math>.
  
 
'''Extension of Feature Space:'''
 
'''Extension of Feature Space:'''
Line 48: Line 79:
 
<math> \{  (x, y, x^2, y^2) \mid (0, -2, 1, 1) \cdot (x, y, x^2, y^2)  = 0 \} </math>
 
<math> \{  (x, y, x^2, y^2) \mid (0, -2, 1, 1) \cdot (x, y, x^2, y^2)  = 0 \} </math>
  
[[Image:lec9_embedd_Old Kiwi.bmp]]
+
[[Image:lec9_embedd_Old Kiwi.jpg]]
  
 
Example 3: 1D example
 
Example 3: 1D example
Line 63: Line 94:
 
This can be expressed as below mathematical form.
 
This can be expressed as below mathematical form.
  
|jinha_pp2|
+
<math>  p(x \mid w_1) \rightarrow p(x, x^2 \mid w_1) </math>
  
.. |jinha_pp2| image:: tex
+
'''Taylor Series:''' If true <math>\vec{g}</math> is analytic
:alt: tex: p(x \mid w_1) \rightarrow p(x, x^2 \mid w_1)
+
e.g <math> g(\vec{X})=\sum_{i=1}^{n}{\lambda}_{i}{e}^{{c}_{i}\vec{x}} </math>
  
.. |g_vc| image:: tex
+
then, <math> g(\vec{x}) </math> = Taylor Series --> infinite polynomial in powers
:alt: tex: \vec{g}
+
 
+
.. |g_vec_eq| image:: tex
+
:alt: tex: g(\vec{X})=\sum_{i=1}^{n}{\lambda}_{i}{e}^{{c}_{i}\vec{x}}
+
 
+
.. |g_x| image:: tex
+
:alt: tex: g(\vec{x})
+
 
+
 
+
**Taylor Series** If true |g_vc| is analytic
+
e.g |g_vec_eq|
+
 
+
then, |g_x| = Taylor Series --> infinite polynomial in powers
+
  
 
because we can approximate with Taylor polynomial of degree n
 
because we can approximate with Taylor polynomial of degree n
 
(i.e as n get higher, approximation gets better)
 
(i.e as n get higher, approximation gets better)
  
 
+
[[Image:poly2_Old Kiwi.jpg]]
.. image:: poly2.gif
+
  
 
However, there are issues on this approach. It gets complex with the increase of the dimension of the [feature vector].
 
However, there are issues on this approach. It gets complex with the increase of the dimension of the [feature vector].
  
If |jinha_x|
+
If <math> \vec{x} = (x_1, x_2, \cdots, x_d) </math>
 
+
.. |jinha_x| image:: tex
+
:alt: tex: \vec{x} = (x_1, x_2, \cdots, x_d)
+
 
+
A degree 2 polynomial in |jinha_x| has |jinha_dim2poly| monomials
+
 
+
.. |jinha_dim2poly| image:: tex
+
:alt: tex: \frac{1}{2} (d+1) (d+2) \approx d^2
+
  
A degree n polynomial in |jinha_x| has |jinha_dn| monomials.
+
A degree 2 polynomial in <math> \vec{x} = (x_1, x_2, \cdots, x_d) </math> has <math> \frac{1}{2} (d+1) (d+2) \approx d^2 </math> monomials
  
.. |jinha_dn| image:: tex
+
A degree n polynomial in <math> \vec{x} = (x_1, x_2, \cdots, x_d) </math> has <math> \approx d^n </math> monomials.
:alt: tex: \approx d^n
+
  
 
'''How to find decision hyperplane given training data'''
 
'''How to find decision hyperplane given training data'''
Line 155: Line 163:
 
Since <math> \vec{c} \cdot \vec{y} </math> is proportional to distance from <math> y_{i} </math> to hyperplane
 
Since <math> \vec{c} \cdot \vec{y} </math> is proportional to distance from <math> y_{i} </math> to hyperplane
  
Previous: [[Lecture 8_Old Kiwi]] Next: [[Lecture 10_Old Kiwi]]
+
[[Category:Lecture Notes]]

Latest revision as of 08:48, 17 January 2013

Lecture 9, ECE662: Decision Theory

Lecture notes for ECE662 Spring 2008, Prof. Boutin.

Other lectures: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28,



Parametric Methods

Two applications:

  • Parametric Density Estimation: Using sample data, we estimate probabilities $ P(\omega_i) $, $ p(\vec{X}|\omega_i) \forall i $ etc using estimation methods like MLE_Old Kiwi and BPE_Old Kiwi. Then from these estimated probabilities (and not true probabilities, which are unknown; only their parametric form was known) we can use Bayes classification rule to build a classifier.
  • Parametric Classifiers_Old Kiwi: We find parametric decision boundaries to approximate true decision boundaries between classes. This is very different approach from approximating the probabilities with their estimates, as in previous method.


Example:

Lecture9 parametric decion boundary Old Kiwi.JPG .. image:: Lecture9_parametric_decion_boundary.JPG

True decision boundary: $ \{\vec{X}|P(\omega_1|\vec{X})-P(\omega_2|\vec{X})=0\} $ can be approximated by $ \{\vec{X}|g(\vec{X})=0\} $. Note that 'g' is not an estimate of the difference in probabilities as in true decision boundary, but it is just an approximate parametric form for the true decision boundary. Choice of 'g' depends on whether it is analytic, and hence easy to handle computationally.

E.g. 'g' can be a degree-2 polynomial, or it can be a degree-1 polynomial (resulting in a linear classifier).

If 'g' is linear, it can be written as $ g(\vec{X})=\vec{V}\cdot \vec{X}+V_0=(\vec{V},V_0)\cdot (\vec{X},1) $ or $ g(1,\vec{X})=\vec{c}\cdot(1,\vec{X}) $.

Extension of Feature Space: This trick justifies the importance of studying linear classifiers even though they do not occur so often in practice directly. Actually, many non-linear classifiers can be seen as linear.

Example 1: Consider g to be a polynomial parametric form of a classifier in $ \mathbb{R}^2 $.

$ g(x,y)=c_0+c_1x+c_2y+c_3xy+c_4{x}^2+c_5{y}^2 $

Here g is not a linear classifier in $ \mathbb{R}^2 $. Let us define $ \tilde{g}:\mathbb{R}^5 \to \mathbb{R} $

$ \tilde{g}(u_1,u_2,u_3,u_4,u_5)=c_0+c_1u_1+c_2u_2+c_3u_3+c_4u_4+c_5u_5 $

where $ u_1=x; u_2=y; u_3=xy; u_4={x}^2; u_5={y}^2 $. Here we have defined a parametric from of the classifier in extended feature space. This form is linear. Hence "Decision boundary defined by 'g' is linear in extended feature space."

Example 2: Consider another polynomial,


$ g(x)={x}^2+{y}^2-2{y} = {x}^2+{(y-1)}^2 -1 $

This is a circle centred at (0,1):

Circle Old Kiwi.jpg

This non-linear function can be transformed into linear function in extended feature space like below.

$ \{ (x, y, x^2, y^2) \mid (0, -2, 1, 1) \cdot (x, y, x^2, y^2) = 0 \} $

Lec9 embedd Old Kiwi.jpg

Example 3: 1D example

Jinha 1D Example01 Old Kiwi.jpg

As we can see from above figure, decision boundary is not linear (region is not connected). From this, we can construct an extended [feature vector] space.

$ x \rightarrow (x, x^2) \in R^2 $

Jinha 1D Example02 Old Kiwi.jpg

This red line (which is a hyperplane in this case) can separate these two classes. This can be expressed as below mathematical form.

$ p(x \mid w_1) \rightarrow p(x, x^2 \mid w_1) $

Taylor Series: If true $ \vec{g} $ is analytic e.g $ g(\vec{X})=\sum_{i=1}^{n}{\lambda}_{i}{e}^{{c}_{i}\vec{x}} $

then, $ g(\vec{x}) $ = Taylor Series --> infinite polynomial in powers

because we can approximate with Taylor polynomial of degree n (i.e as n get higher, approximation gets better)

Poly2 Old Kiwi.jpg

However, there are issues on this approach. It gets complex with the increase of the dimension of the [feature vector].

If $ \vec{x} = (x_1, x_2, \cdots, x_d) $

A degree 2 polynomial in $ \vec{x} = (x_1, x_2, \cdots, x_d) $ has $ \frac{1}{2} (d+1) (d+2) \approx d^2 $ monomials

A degree n polynomial in $ \vec{x} = (x_1, x_2, \cdots, x_d) $ has $ \approx d^n $ monomials.

How to find decision hyperplane given training data

In the best of all cases, data is "linearly separable", meaning there exists a vector c such that for all y training data: c * y > 0 if y belongs to class w1. (where * is a dot product) c * y < 0 of u belongs to class w2.

Trick: replace all y's in the training data belonging to class w2 by -y, then look for vector c such that c * y > 0 for all y.

Example: 1D feature space

x->(1,x)=:y

Lecture9 1D yi -yi Old Kiwi.JPG

First component is +1 for all of class 1 training data

First component is -1 for all of class 2 training data

Lecture9 1D yi -yi 2 Old Kiwi.JPG

Observe: c is not unique

To make $ \vec{c} $ unique, introduce 'margin' b>0 and ask that c be the minimum length vector such that $ \vec{c} \cdot y_{i} \geq b $, for all i

Geometric Interpretation of margin

If $ \vec{c} \cdot y_{i} = b $, then $ \vec{c} $ belongs to hyperplane with normal vector $ y_{i0} $ which is at distance $ \displaystyle \frac{b} {\displaystyle ||y_{i0} ||} $ from the origin.


$ \vec{c} \cdot y_{i} \geq b $, for all i means that $ \vec{c} $ is always at least a distance $ \displaystyle \frac{b} {\displaystyle ||y_{i0} ||} $ from the origin for all i

In general, it is impossible to satisfy $ \vec{c} \cdot y_{i} \geq 0 $, for all i

Perhaps, we could try to minimize number of misclassfied training samples

Let $ J(\vec{c})=\{ y_{i} | \vec{c} \cdot y_{i} \leq 0 \} $ -> Too hard to minimize.

Perceptron_Old Kiwi Criterion function

$ J_{p}(\vec{c})=\displaystyle \sum_{y_i, missclassified} (- \vec{c} \cdot y_{i} ) $

Lec9 percept Old Kiwi.JPG

Measure how 'far away' you are from bias right. distance from |y_i| to hyperplane defined by |vec_c| is $ ||\vec{y_i} \cdot \frac{\vec{c}}{||\vec{c}||} ||=\frac{1}{|| \vec{c} ||} \vec{y_i} \cdot \vec{c} $

Since $ \vec{c} \cdot \vec{y} $ is proportional to distance from $ y_{i} $ to hyperplane

Alumni Liaison

Abstract algebra continues the conceptual developments of linear algebra, on an even grander scale.

Dr. Paul Garrett