SlideShare a Scribd company logo
Knowledge Extraction from
Support Vector Machines:
A Fuzzy Logic Approach
“Certain class of SVMs is mathematically
equivalent to FARB”
What is SVM
What does it do?
Learns a hyper plane to classify data into
2 classes.
What is a hyperplane?
A hyperplane is a function like the equation for a line,
𝑦 = 𝑚𝑥 + 𝑏
In fact, for a simple classification task with just 2
features, the hyperplane can be a line.
SVM finds the optimal solution.
Support Vector Machine
SVM attempts to maximize
the margin, so that the
hyperplane is just as far away
from red ball as the blue ball.
In this way, it decreases the
chance of misclassification.
More Formally
Input:
set of (input, output) training pair samples.
Output:
set of weights w (or 𝑤𝑖), one for each feature, whose
linear combination predicts the value of y.
We use the optimization of maximizing the
margin (‘street width’) to reduce the number
of weights that are nonzero to just a few that
correspond to the important features that
‘matter’ in deciding the separating
line(hyperplane)…these nonzero weights
correspond to the support vectors (because
they ‘support’ the separating hyperplane)
The optimization problem
minimize 𝑓(𝑤) ≡ (1/2) ∥ 𝒘 ∥2
subject to 𝑔 𝑤, 𝑏 ≡ −𝑦𝑖 𝒘 ⋅ 𝒙 + 𝑏 + 1 ≤ 0, 𝑖 = 1 … 𝑚
we use Lagrange multipliers to get this
problem into a form that can be solved analytically
What if things get more complicated?
Throw the balls in the air. While the balls are
in the air and thrown up in just the right way,
you use a large sheet of paper to divide the
balls in the air.
mapping data to a high dimensional
space
Kernel
polynomial: (𝒙𝒊 ⋅ 𝒙𝒋 + 𝑐) 𝑝
Gaussian radial basis function: exp(−∥ 𝒙𝒊– 𝒙𝒋 ∥2
/2𝜎2
)
SVM does its thing, maps them into a higher dimension
and then finds the hyperplane to separate the classes.
Where does SVM get its name from?
• The decision function is fully specified by a (usually very small) subset
of training samples, the support vectors.
• Support vectors are the data points that lie closest to the decision
surface (or hyperplane)
• They are the data points most difficult to classify
• They have direct bearing on the optimum location of the decision
surface
• they ‘support’ the separating hyperplane
Knowledge Extraction
Extracting the knowledge learned by a black–box classifier and
representing it in a comprehensible form
Knowledge Extraction
Benefits :
• Validation
• Feature extraction
• Knowledge refinement and improvement
• Knowledge acquisition for symbolic AI systems
• Scientific discovery
Knowledge Extraction Rule Extraction
• Methods for RE from ANNs have been classified into three categories:
Decompositional
Pedagogical
Eclectic
decompositional approach for KE
SVM :The IO mapping of the trained SVM f : Rn → {−1, 1} is given by
decompositional approach for KE
“Certain class of SVMs is mathematically
equivalent to FARB”
What is FARB !?
Let’s take an example first !
Example:
Input: q ∈ R ,
Output: O ∈ R ,
And: a0, a1, k ∈ R, with k > 0.
Rules:
R1: If q is equal to k Then O = a0 + a1,
R2: If q is equal to −k Then O = a0 − a1,
- Linguistic terms: equal to k , equal to –k
- To express fuzziness, Gaussian membership function is used:
These Function Satisfy:
Applying Singleton fuzzifier and Centre of Gravity defuzzifier yields:
But, What does this Output mean !??
Take a deeper look !
It is a feedforward ANN with a single neuron, employing the activation
function tanh() !
So: this FRB is equivalent to ANN
This FRB , in particular, satisfy the definition of FARB, which is:
To get the same output, apply the same steps as in the example, which
is:
But how this output is any close to the one in the example ?!?
And many other MFs satisfy this output, given specific values of z,u,v,r
and g. Such as Logistic function and others.
Apply: zi = ui = 1,
vi = ri = 0,
and gi(x) = tanh(x).
Result
Kolman and Margaliot: Every standard ANN has a corresponding FARB.
There’s a transformation T:
This work extend that to: Certain class of SVMs satisfy the
transformation P:
“Certain class of SVMs is mathematically
equivalent to FARB”
The SVM-FARB Equivalence
*
1
( ) * ( , )
Nsv
i
i i
i
h x b y K x s

 
(2)
0
1 1
( ) ( )
m m
i i i i i i i i
i i
O q a ra z a g u q v
 
    
(8)
Theorem 2. (SVM-FARB equivalence)
condition
Find FARB with:
So these conditions would hold
0, , , ,i i i i
m Nsv
q a a   

*
0
1
*
,
,
( ) ( , )
m
i i
i
i i i i
i
i i i i
a ra b
z a y
g u q v K x s


 

 

(15)
Pause and Think
• Let’s say we have a FARB
• How many rules have we got?
1 1
0 1
...
...
m m
m
If q is and and q is
Then O a a a
  
   
(7)
Famous SVM Kernels
( , ) , (linear kernel)T
K x y x y
( , ) (1 / ) , , , (polynomial kernel)T d
K x y x y c c d   
( , ) tanh( ), 0, 0, (MLP kernel)T
K x y x y      
2 2
ˆ ˆ( , ) exp( / (2 )), , (RBF or Gaussian kernel)K x y x y     
Corollary 1 {MLP kernel}
These parameters will satisfy (15) conditions
( ) ( )
tanh(( ) )
( ) ( ) 2
k k
k k
q q
q k
q q
  
 
 
 

 

,
,
2 , /
i i
T i
i
i i
k k
i i
q x s
k
   
   
   

 
  
* *
1
( ) tanh( )
Nsv
T i
i i
i
h x y x s b  

   (17)
0
1 1
( ) ( )
m m
i i i i i i i i
i i
O q a ra z a g u q v
 
    
* *
0
1, 0,
2 ,
2 ,
( ) tanh( )
,
i i
i i
i i i
i
i i i
z r
u
v k
g x x
a b a y
 
 

 
 
  

 
Pause and Think
appear in the FARB if-part
What could this mean?
iq
cos
cos ; and are normalized
T i
i
i
i
i
i
q x s
q x s
q x s





Corollary 2 {MPL Kernel}
2
* *
2
1
( ) exp( )
ˆ2
Nsv
i i
i
x y
h x y b


   (18)
These parameters will
satisfy (15) conditions
0
1 1
( ) ( )
m m
i i i i i i i i
i i
O q a ra z a g u q v
 
    
2
2
( ) ( ) ( )
2exp( ) 1
( ) ( ) 2
k k
k k
q q q k
q q
 
  
 
 
 
  

0 0
,
,
ˆ, 0
T i
i
i i
q x s
k
   
 
   
 
 
 
2
2
* *
0 1
*
2, 1
1 2 ,
0,
( ) exp( ),
/ 2,
/ 2
i i
i
i
i
Nsv
i ii
i i i
z r
u
v
g x x
a b y
a y




  


 
 


Experimentation{Iris data set}
• 150 examples
• sepal length
• sepal width
• length
• petal width
• 3 classes
• Setosa
• Versicolor
• Virginica
SVM
Results {SVM1}
Results {SVM2}
Knowledge extraction from support vector machines
Knowledge extraction from support vector machines

More Related Content

What's hot

Optics Fourier Transform I
Optics Fourier Transform IOptics Fourier Transform I
Optics Fourier Transform Idiarmseven
 
Presentation on fourier transformation
Presentation on fourier transformationPresentation on fourier transformation
Presentation on fourier transformation
Wasim Shah
 
Fourier series and applications of fourier transform
Fourier series and applications of fourier transformFourier series and applications of fourier transform
Fourier series and applications of fourier transform
Krishna Jangid
 
06. string matching
06. string matching06. string matching
06. string matching
Onkar Nath Sharma
 
Discrete Fourier Series | Discrete Fourier Transform | Discrete Time Fourier ...
Discrete Fourier Series | Discrete Fourier Transform | Discrete Time Fourier ...Discrete Fourier Series | Discrete Fourier Transform | Discrete Time Fourier ...
Discrete Fourier Series | Discrete Fourier Transform | Discrete Time Fourier ...
Mehran University Of Engineering and Technology, Pakistan
 
Fourier series and fourier integral
Fourier series and fourier integralFourier series and fourier integral
Fourier series and fourier integral
ashuuhsaqwe
 
Half range sine cosine fourier series
Half range sine cosine fourier seriesHalf range sine cosine fourier series
Half range sine cosine fourier seriesHardik Parmar
 
Fuzzy inference
Fuzzy inferenceFuzzy inference
Fuzzy inference
swati singh
 
Fuzzy rules and fuzzy reasoning
Fuzzy rules and fuzzy reasoningFuzzy rules and fuzzy reasoning
Fuzzy rules and fuzzy reasoning
Veni7
 
Half range sine and cosine series
Half range sine and cosine seriesHalf range sine and cosine series
Half range sine and cosine series
Chandan S
 
aem : Fourier series of Even and Odd Function
aem :  Fourier series of Even and Odd Functionaem :  Fourier series of Even and Odd Function
aem : Fourier series of Even and Odd Function
Sukhvinder Singh
 
1531 fourier series- integrals and trans
1531 fourier series- integrals and trans1531 fourier series- integrals and trans
1531 fourier series- integrals and trans
Dr Fereidoun Dejahang
 
Applied Calculus Chapter 2 vector valued function
Applied Calculus Chapter  2 vector valued functionApplied Calculus Chapter  2 vector valued function
Applied Calculus Chapter 2 vector valued function
J C
 
Fourier series
Fourier seriesFourier series
Fourier series
Shiv Prasad Gupta
 
07 periodic functions and fourier series
07 periodic functions and fourier series07 periodic functions and fourier series
07 periodic functions and fourier series
Krishna Gali
 
the fourier series
the fourier seriesthe fourier series
the fourier seriessafi al amu
 
Fourier series and it's examples
Fourier series and it's examplesFourier series and it's examples
Fourier series and it's examples
ZorawarJaat
 
presentation_final
presentation_finalpresentation_final
presentation_finalSushanta Roy
 

What's hot (20)

Optics Fourier Transform I
Optics Fourier Transform IOptics Fourier Transform I
Optics Fourier Transform I
 
Presentation on fourier transformation
Presentation on fourier transformationPresentation on fourier transformation
Presentation on fourier transformation
 
Fourier series and applications of fourier transform
Fourier series and applications of fourier transformFourier series and applications of fourier transform
Fourier series and applications of fourier transform
 
06. string matching
06. string matching06. string matching
06. string matching
 
Discrete Fourier Series | Discrete Fourier Transform | Discrete Time Fourier ...
Discrete Fourier Series | Discrete Fourier Transform | Discrete Time Fourier ...Discrete Fourier Series | Discrete Fourier Transform | Discrete Time Fourier ...
Discrete Fourier Series | Discrete Fourier Transform | Discrete Time Fourier ...
 
Fourier series and fourier integral
Fourier series and fourier integralFourier series and fourier integral
Fourier series and fourier integral
 
Half range sine cosine fourier series
Half range sine cosine fourier seriesHalf range sine cosine fourier series
Half range sine cosine fourier series
 
Fourier transform
Fourier transformFourier transform
Fourier transform
 
Fuzzy inference
Fuzzy inferenceFuzzy inference
Fuzzy inference
 
Fuzzy rules and fuzzy reasoning
Fuzzy rules and fuzzy reasoningFuzzy rules and fuzzy reasoning
Fuzzy rules and fuzzy reasoning
 
Half range sine and cosine series
Half range sine and cosine seriesHalf range sine and cosine series
Half range sine and cosine series
 
aem : Fourier series of Even and Odd Function
aem :  Fourier series of Even and Odd Functionaem :  Fourier series of Even and Odd Function
aem : Fourier series of Even and Odd Function
 
1531 fourier series- integrals and trans
1531 fourier series- integrals and trans1531 fourier series- integrals and trans
1531 fourier series- integrals and trans
 
Calc 5.8a
Calc 5.8aCalc 5.8a
Calc 5.8a
 
Applied Calculus Chapter 2 vector valued function
Applied Calculus Chapter  2 vector valued functionApplied Calculus Chapter  2 vector valued function
Applied Calculus Chapter 2 vector valued function
 
Fourier series
Fourier seriesFourier series
Fourier series
 
07 periodic functions and fourier series
07 periodic functions and fourier series07 periodic functions and fourier series
07 periodic functions and fourier series
 
the fourier series
the fourier seriesthe fourier series
the fourier series
 
Fourier series and it's examples
Fourier series and it's examplesFourier series and it's examples
Fourier series and it's examples
 
presentation_final
presentation_finalpresentation_final
presentation_final
 

Similar to Knowledge extraction from support vector machines

An evolutionary method for constructing complex SVM kernels
An evolutionary method for constructing complex SVM kernelsAn evolutionary method for constructing complex SVM kernels
An evolutionary method for constructing complex SVM kernels
infopapers
 
The world of loss function
The world of loss functionThe world of loss function
The world of loss function
홍배 김
 
Anomaly detection using deep one class classifier
Anomaly detection using deep one class classifierAnomaly detection using deep one class classifier
Anomaly detection using deep one class classifier
홍배 김
 
Optimization Of Fuzzy Bexa Using Nm
Optimization Of Fuzzy Bexa Using NmOptimization Of Fuzzy Bexa Using Nm
Optimization Of Fuzzy Bexa Using Nm
Ashish Khetan
 
Support Vector Machine.pptx
Support Vector Machine.pptxSupport Vector Machine.pptx
Support Vector Machine.pptx
HarishNayak44
 
Introduction to Support Vector Machines
Introduction to Support Vector MachinesIntroduction to Support Vector Machines
Introduction to Support Vector Machines
Silicon Mentor
 
Module 1 ppt class.pptx
Module 1 ppt class.pptxModule 1 ppt class.pptx
Module 1 ppt class.pptx
VivekNaik55
 
Module ppt class.pptx
Module ppt class.pptxModule ppt class.pptx
Module ppt class.pptx
VivekNaik71
 
Dycops2019
Dycops2019 Dycops2019
Dycops2019
Jéssyca Bessa
 
Hmm and neural networks
Hmm and neural networksHmm and neural networks
Hmm and neural networks
Janani Ramasamy
 
Support vector machine
Support vector machineSupport vector machine
Support vector machine
Rishabh Gupta
 
Quality Python Homework Help
Quality Python Homework HelpQuality Python Homework Help
Quality Python Homework Help
Python Homework Help
 
Programmable PN Sequence Generators
Programmable PN Sequence GeneratorsProgrammable PN Sequence Generators
Programmable PN Sequence Generators
Rajesh Singh
 
SVM (Support Vector Machine & Kernel)
SVM (Support Vector Machine & Kernel)SVM (Support Vector Machine & Kernel)
SVM (Support Vector Machine & Kernel)
SEMINARGROOT
 
Non linearequationsmatlab
Non linearequationsmatlabNon linearequationsmatlab
Non linearequationsmatlab
sheetslibrary
 
Non linearequationsmatlab
Non linearequationsmatlabNon linearequationsmatlab
Non linearequationsmatlabZunAib Ali
 
Solution of non-linear equations
Solution of non-linear equationsSolution of non-linear equations
Solution of non-linear equationsZunAib Ali
 
Machine Learning Algorithms (Part 1)
Machine Learning Algorithms (Part 1)Machine Learning Algorithms (Part 1)
Machine Learning Algorithms (Part 1)
Zihui Li
 
CS221: HMM and Particle Filters
CS221: HMM and Particle FiltersCS221: HMM and Particle Filters
CS221: HMM and Particle Filterszukun
 

Similar to Knowledge extraction from support vector machines (20)

An evolutionary method for constructing complex SVM kernels
An evolutionary method for constructing complex SVM kernelsAn evolutionary method for constructing complex SVM kernels
An evolutionary method for constructing complex SVM kernels
 
The world of loss function
The world of loss functionThe world of loss function
The world of loss function
 
Anomaly detection using deep one class classifier
Anomaly detection using deep one class classifierAnomaly detection using deep one class classifier
Anomaly detection using deep one class classifier
 
Optimization Of Fuzzy Bexa Using Nm
Optimization Of Fuzzy Bexa Using NmOptimization Of Fuzzy Bexa Using Nm
Optimization Of Fuzzy Bexa Using Nm
 
Support Vector Machine.pptx
Support Vector Machine.pptxSupport Vector Machine.pptx
Support Vector Machine.pptx
 
Zoooooohaib
ZoooooohaibZoooooohaib
Zoooooohaib
 
Introduction to Support Vector Machines
Introduction to Support Vector MachinesIntroduction to Support Vector Machines
Introduction to Support Vector Machines
 
Module 1 ppt class.pptx
Module 1 ppt class.pptxModule 1 ppt class.pptx
Module 1 ppt class.pptx
 
Module ppt class.pptx
Module ppt class.pptxModule ppt class.pptx
Module ppt class.pptx
 
Dycops2019
Dycops2019 Dycops2019
Dycops2019
 
Hmm and neural networks
Hmm and neural networksHmm and neural networks
Hmm and neural networks
 
Support vector machine
Support vector machineSupport vector machine
Support vector machine
 
Quality Python Homework Help
Quality Python Homework HelpQuality Python Homework Help
Quality Python Homework Help
 
Programmable PN Sequence Generators
Programmable PN Sequence GeneratorsProgrammable PN Sequence Generators
Programmable PN Sequence Generators
 
SVM (Support Vector Machine & Kernel)
SVM (Support Vector Machine & Kernel)SVM (Support Vector Machine & Kernel)
SVM (Support Vector Machine & Kernel)
 
Non linearequationsmatlab
Non linearequationsmatlabNon linearequationsmatlab
Non linearequationsmatlab
 
Non linearequationsmatlab
Non linearequationsmatlabNon linearequationsmatlab
Non linearequationsmatlab
 
Solution of non-linear equations
Solution of non-linear equationsSolution of non-linear equations
Solution of non-linear equations
 
Machine Learning Algorithms (Part 1)
Machine Learning Algorithms (Part 1)Machine Learning Algorithms (Part 1)
Machine Learning Algorithms (Part 1)
 
CS221: HMM and Particle Filters
CS221: HMM and Particle FiltersCS221: HMM and Particle Filters
CS221: HMM and Particle Filters
 

Recently uploaded

一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
zwunae
 
Governing Equations for Fundamental Aerodynamics_Anderson2010.pdf
Governing Equations for Fundamental Aerodynamics_Anderson2010.pdfGoverning Equations for Fundamental Aerodynamics_Anderson2010.pdf
Governing Equations for Fundamental Aerodynamics_Anderson2010.pdf
WENKENLI1
 
一比一原版(SFU毕业证)西蒙菲莎大学毕业证成绩单如何办理
一比一原版(SFU毕业证)西蒙菲莎大学毕业证成绩单如何办理一比一原版(SFU毕业证)西蒙菲莎大学毕业证成绩单如何办理
一比一原版(SFU毕业证)西蒙菲莎大学毕业证成绩单如何办理
bakpo1
 
Understanding Inductive Bias in Machine Learning
Understanding Inductive Bias in Machine LearningUnderstanding Inductive Bias in Machine Learning
Understanding Inductive Bias in Machine Learning
SUTEJAS
 
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
obonagu
 
Water Industry Process Automation and Control Monthly - May 2024.pdf
Water Industry Process Automation and Control Monthly - May 2024.pdfWater Industry Process Automation and Control Monthly - May 2024.pdf
Water Industry Process Automation and Control Monthly - May 2024.pdf
Water Industry Process Automation & Control
 
Tutorial for 16S rRNA Gene Analysis with QIIME2.pdf
Tutorial for 16S rRNA Gene Analysis with QIIME2.pdfTutorial for 16S rRNA Gene Analysis with QIIME2.pdf
Tutorial for 16S rRNA Gene Analysis with QIIME2.pdf
aqil azizi
 
Final project report on grocery store management system..pdf
Final project report on grocery store management system..pdfFinal project report on grocery store management system..pdf
Final project report on grocery store management system..pdf
Kamal Acharya
 
Harnessing WebAssembly for Real-time Stateless Streaming Pipelines
Harnessing WebAssembly for Real-time Stateless Streaming PipelinesHarnessing WebAssembly for Real-time Stateless Streaming Pipelines
Harnessing WebAssembly for Real-time Stateless Streaming Pipelines
Christina Lin
 
Industrial Training at Shahjalal Fertilizer Company Limited (SFCL)
Industrial Training at Shahjalal Fertilizer Company Limited (SFCL)Industrial Training at Shahjalal Fertilizer Company Limited (SFCL)
Industrial Training at Shahjalal Fertilizer Company Limited (SFCL)
MdTanvirMahtab2
 
RAT: Retrieval Augmented Thoughts Elicit Context-Aware Reasoning in Long-Hori...
RAT: Retrieval Augmented Thoughts Elicit Context-Aware Reasoning in Long-Hori...RAT: Retrieval Augmented Thoughts Elicit Context-Aware Reasoning in Long-Hori...
RAT: Retrieval Augmented Thoughts Elicit Context-Aware Reasoning in Long-Hori...
thanhdowork
 
PPT on GRP pipes manufacturing and testing
PPT on GRP pipes manufacturing and testingPPT on GRP pipes manufacturing and testing
PPT on GRP pipes manufacturing and testing
anoopmanoharan2
 
Hybrid optimization of pumped hydro system and solar- Engr. Abdul-Azeez.pdf
Hybrid optimization of pumped hydro system and solar- Engr. Abdul-Azeez.pdfHybrid optimization of pumped hydro system and solar- Engr. Abdul-Azeez.pdf
Hybrid optimization of pumped hydro system and solar- Engr. Abdul-Azeez.pdf
fxintegritypublishin
 
NO1 Uk best vashikaran specialist in delhi vashikaran baba near me online vas...
NO1 Uk best vashikaran specialist in delhi vashikaran baba near me online vas...NO1 Uk best vashikaran specialist in delhi vashikaran baba near me online vas...
NO1 Uk best vashikaran specialist in delhi vashikaran baba near me online vas...
Amil Baba Dawood bangali
 
14 Template Contractual Notice - EOT Application
14 Template Contractual Notice - EOT Application14 Template Contractual Notice - EOT Application
14 Template Contractual Notice - EOT Application
SyedAbiiAzazi1
 
MCQ Soil mechanics questions (Soil shear strength).pdf
MCQ Soil mechanics questions (Soil shear strength).pdfMCQ Soil mechanics questions (Soil shear strength).pdf
MCQ Soil mechanics questions (Soil shear strength).pdf
Osamah Alsalih
 
Planning Of Procurement o different goods and services
Planning Of Procurement o different goods and servicesPlanning Of Procurement o different goods and services
Planning Of Procurement o different goods and services
JoytuBarua2
 
Hierarchical Digital Twin of a Naval Power System
Hierarchical Digital Twin of a Naval Power SystemHierarchical Digital Twin of a Naval Power System
Hierarchical Digital Twin of a Naval Power System
Kerry Sado
 
Student information management system project report ii.pdf
Student information management system project report ii.pdfStudent information management system project report ii.pdf
Student information management system project report ii.pdf
Kamal Acharya
 
NUMERICAL SIMULATIONS OF HEAT AND MASS TRANSFER IN CONDENSING HEAT EXCHANGERS...
NUMERICAL SIMULATIONS OF HEAT AND MASS TRANSFER IN CONDENSING HEAT EXCHANGERS...NUMERICAL SIMULATIONS OF HEAT AND MASS TRANSFER IN CONDENSING HEAT EXCHANGERS...
NUMERICAL SIMULATIONS OF HEAT AND MASS TRANSFER IN CONDENSING HEAT EXCHANGERS...
ssuser7dcef0
 

Recently uploaded (20)

一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
一比一原版(IIT毕业证)伊利诺伊理工大学毕业证成绩单专业办理
 
Governing Equations for Fundamental Aerodynamics_Anderson2010.pdf
Governing Equations for Fundamental Aerodynamics_Anderson2010.pdfGoverning Equations for Fundamental Aerodynamics_Anderson2010.pdf
Governing Equations for Fundamental Aerodynamics_Anderson2010.pdf
 
一比一原版(SFU毕业证)西蒙菲莎大学毕业证成绩单如何办理
一比一原版(SFU毕业证)西蒙菲莎大学毕业证成绩单如何办理一比一原版(SFU毕业证)西蒙菲莎大学毕业证成绩单如何办理
一比一原版(SFU毕业证)西蒙菲莎大学毕业证成绩单如何办理
 
Understanding Inductive Bias in Machine Learning
Understanding Inductive Bias in Machine LearningUnderstanding Inductive Bias in Machine Learning
Understanding Inductive Bias in Machine Learning
 
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
在线办理(ANU毕业证书)澳洲国立大学毕业证录取通知书一模一样
 
Water Industry Process Automation and Control Monthly - May 2024.pdf
Water Industry Process Automation and Control Monthly - May 2024.pdfWater Industry Process Automation and Control Monthly - May 2024.pdf
Water Industry Process Automation and Control Monthly - May 2024.pdf
 
Tutorial for 16S rRNA Gene Analysis with QIIME2.pdf
Tutorial for 16S rRNA Gene Analysis with QIIME2.pdfTutorial for 16S rRNA Gene Analysis with QIIME2.pdf
Tutorial for 16S rRNA Gene Analysis with QIIME2.pdf
 
Final project report on grocery store management system..pdf
Final project report on grocery store management system..pdfFinal project report on grocery store management system..pdf
Final project report on grocery store management system..pdf
 
Harnessing WebAssembly for Real-time Stateless Streaming Pipelines
Harnessing WebAssembly for Real-time Stateless Streaming PipelinesHarnessing WebAssembly for Real-time Stateless Streaming Pipelines
Harnessing WebAssembly for Real-time Stateless Streaming Pipelines
 
Industrial Training at Shahjalal Fertilizer Company Limited (SFCL)
Industrial Training at Shahjalal Fertilizer Company Limited (SFCL)Industrial Training at Shahjalal Fertilizer Company Limited (SFCL)
Industrial Training at Shahjalal Fertilizer Company Limited (SFCL)
 
RAT: Retrieval Augmented Thoughts Elicit Context-Aware Reasoning in Long-Hori...
RAT: Retrieval Augmented Thoughts Elicit Context-Aware Reasoning in Long-Hori...RAT: Retrieval Augmented Thoughts Elicit Context-Aware Reasoning in Long-Hori...
RAT: Retrieval Augmented Thoughts Elicit Context-Aware Reasoning in Long-Hori...
 
PPT on GRP pipes manufacturing and testing
PPT on GRP pipes manufacturing and testingPPT on GRP pipes manufacturing and testing
PPT on GRP pipes manufacturing and testing
 
Hybrid optimization of pumped hydro system and solar- Engr. Abdul-Azeez.pdf
Hybrid optimization of pumped hydro system and solar- Engr. Abdul-Azeez.pdfHybrid optimization of pumped hydro system and solar- Engr. Abdul-Azeez.pdf
Hybrid optimization of pumped hydro system and solar- Engr. Abdul-Azeez.pdf
 
NO1 Uk best vashikaran specialist in delhi vashikaran baba near me online vas...
NO1 Uk best vashikaran specialist in delhi vashikaran baba near me online vas...NO1 Uk best vashikaran specialist in delhi vashikaran baba near me online vas...
NO1 Uk best vashikaran specialist in delhi vashikaran baba near me online vas...
 
14 Template Contractual Notice - EOT Application
14 Template Contractual Notice - EOT Application14 Template Contractual Notice - EOT Application
14 Template Contractual Notice - EOT Application
 
MCQ Soil mechanics questions (Soil shear strength).pdf
MCQ Soil mechanics questions (Soil shear strength).pdfMCQ Soil mechanics questions (Soil shear strength).pdf
MCQ Soil mechanics questions (Soil shear strength).pdf
 
Planning Of Procurement o different goods and services
Planning Of Procurement o different goods and servicesPlanning Of Procurement o different goods and services
Planning Of Procurement o different goods and services
 
Hierarchical Digital Twin of a Naval Power System
Hierarchical Digital Twin of a Naval Power SystemHierarchical Digital Twin of a Naval Power System
Hierarchical Digital Twin of a Naval Power System
 
Student information management system project report ii.pdf
Student information management system project report ii.pdfStudent information management system project report ii.pdf
Student information management system project report ii.pdf
 
NUMERICAL SIMULATIONS OF HEAT AND MASS TRANSFER IN CONDENSING HEAT EXCHANGERS...
NUMERICAL SIMULATIONS OF HEAT AND MASS TRANSFER IN CONDENSING HEAT EXCHANGERS...NUMERICAL SIMULATIONS OF HEAT AND MASS TRANSFER IN CONDENSING HEAT EXCHANGERS...
NUMERICAL SIMULATIONS OF HEAT AND MASS TRANSFER IN CONDENSING HEAT EXCHANGERS...
 

Knowledge extraction from support vector machines

  • 1. Knowledge Extraction from Support Vector Machines: A Fuzzy Logic Approach
  • 2. “Certain class of SVMs is mathematically equivalent to FARB”
  • 4. What does it do? Learns a hyper plane to classify data into 2 classes.
  • 5. What is a hyperplane? A hyperplane is a function like the equation for a line, 𝑦 = 𝑚𝑥 + 𝑏 In fact, for a simple classification task with just 2 features, the hyperplane can be a line.
  • 6. SVM finds the optimal solution.
  • 7. Support Vector Machine SVM attempts to maximize the margin, so that the hyperplane is just as far away from red ball as the blue ball. In this way, it decreases the chance of misclassification.
  • 8. More Formally Input: set of (input, output) training pair samples. Output: set of weights w (or 𝑤𝑖), one for each feature, whose linear combination predicts the value of y.
  • 9. We use the optimization of maximizing the margin (‘street width’) to reduce the number of weights that are nonzero to just a few that correspond to the important features that ‘matter’ in deciding the separating line(hyperplane)…these nonzero weights correspond to the support vectors (because they ‘support’ the separating hyperplane)
  • 10. The optimization problem minimize 𝑓(𝑤) ≡ (1/2) ∥ 𝒘 ∥2 subject to 𝑔 𝑤, 𝑏 ≡ −𝑦𝑖 𝒘 ⋅ 𝒙 + 𝑏 + 1 ≤ 0, 𝑖 = 1 … 𝑚 we use Lagrange multipliers to get this problem into a form that can be solved analytically
  • 11. What if things get more complicated?
  • 12. Throw the balls in the air. While the balls are in the air and thrown up in just the right way, you use a large sheet of paper to divide the balls in the air. mapping data to a high dimensional space
  • 13. Kernel polynomial: (𝒙𝒊 ⋅ 𝒙𝒋 + 𝑐) 𝑝 Gaussian radial basis function: exp(−∥ 𝒙𝒊– 𝒙𝒋 ∥2 /2𝜎2 ) SVM does its thing, maps them into a higher dimension and then finds the hyperplane to separate the classes.
  • 14.
  • 15. Where does SVM get its name from? • The decision function is fully specified by a (usually very small) subset of training samples, the support vectors. • Support vectors are the data points that lie closest to the decision surface (or hyperplane) • They are the data points most difficult to classify • They have direct bearing on the optimum location of the decision surface • they ‘support’ the separating hyperplane
  • 16. Knowledge Extraction Extracting the knowledge learned by a black–box classifier and representing it in a comprehensible form
  • 17. Knowledge Extraction Benefits : • Validation • Feature extraction • Knowledge refinement and improvement • Knowledge acquisition for symbolic AI systems • Scientific discovery
  • 18. Knowledge Extraction Rule Extraction • Methods for RE from ANNs have been classified into three categories: Decompositional Pedagogical Eclectic
  • 19. decompositional approach for KE SVM :The IO mapping of the trained SVM f : Rn → {−1, 1} is given by
  • 21.
  • 22.
  • 23. “Certain class of SVMs is mathematically equivalent to FARB”
  • 24. What is FARB !? Let’s take an example first !
  • 25. Example: Input: q ∈ R , Output: O ∈ R , And: a0, a1, k ∈ R, with k > 0. Rules: R1: If q is equal to k Then O = a0 + a1, R2: If q is equal to −k Then O = a0 − a1,
  • 26. - Linguistic terms: equal to k , equal to –k - To express fuzziness, Gaussian membership function is used:
  • 28. Applying Singleton fuzzifier and Centre of Gravity defuzzifier yields: But, What does this Output mean !??
  • 29. Take a deeper look ! It is a feedforward ANN with a single neuron, employing the activation function tanh() ! So: this FRB is equivalent to ANN
  • 30. This FRB , in particular, satisfy the definition of FARB, which is:
  • 31.
  • 32.
  • 33. To get the same output, apply the same steps as in the example, which is: But how this output is any close to the one in the example ?!?
  • 34. And many other MFs satisfy this output, given specific values of z,u,v,r and g. Such as Logistic function and others. Apply: zi = ui = 1, vi = ri = 0, and gi(x) = tanh(x).
  • 35. Result Kolman and Margaliot: Every standard ANN has a corresponding FARB. There’s a transformation T: This work extend that to: Certain class of SVMs satisfy the transformation P:
  • 36. “Certain class of SVMs is mathematically equivalent to FARB”
  • 37. The SVM-FARB Equivalence * 1 ( ) * ( , ) Nsv i i i i h x b y K x s    (2) 0 1 1 ( ) ( ) m m i i i i i i i i i i O q a ra z a g u q v        (8)
  • 38. Theorem 2. (SVM-FARB equivalence) condition Find FARB with: So these conditions would hold 0, , , ,i i i i m Nsv q a a     * 0 1 * , , ( ) ( , ) m i i i i i i i i i i i i a ra b z a y g u q v K x s         (15)
  • 39. Pause and Think • Let’s say we have a FARB • How many rules have we got? 1 1 0 1 ... ... m m m If q is and and q is Then O a a a        (7)
  • 40. Famous SVM Kernels ( , ) , (linear kernel)T K x y x y ( , ) (1 / ) , , , (polynomial kernel)T d K x y x y c c d    ( , ) tanh( ), 0, 0, (MLP kernel)T K x y x y       2 2 ˆ ˆ( , ) exp( / (2 )), , (RBF or Gaussian kernel)K x y x y     
  • 41. Corollary 1 {MLP kernel} These parameters will satisfy (15) conditions ( ) ( ) tanh(( ) ) ( ) ( ) 2 k k k k q q q k q q              , , 2 , / i i T i i i i k k i i q x s k                   * * 1 ( ) tanh( ) Nsv T i i i i h x y x s b       (17) 0 1 1 ( ) ( ) m m i i i i i i i i i i O q a ra z a g u q v        * * 0 1, 0, 2 , 2 , ( ) tanh( ) , i i i i i i i i i i i z r u v k g x x a b a y               
  • 42. Pause and Think appear in the FARB if-part What could this mean? iq cos cos ; and are normalized T i i i i i i q x s q x s q x s     
  • 43. Corollary 2 {MPL Kernel} 2 * * 2 1 ( ) exp( ) ˆ2 Nsv i i i x y h x y b      (18) These parameters will satisfy (15) conditions 0 1 1 ( ) ( ) m m i i i i i i i i i i O q a ra z a g u q v        2 2 ( ) ( ) ( ) 2exp( ) 1 ( ) ( ) 2 k k k k q q q k q q                0 0 , , ˆ, 0 T i i i i q x s k                 2 2 * * 0 1 * 2, 1 1 2 , 0, ( ) exp( ), / 2, / 2 i i i i i Nsv i ii i i i z r u v g x x a b y a y               
  • 44. Experimentation{Iris data set} • 150 examples • sepal length • sepal width • length • petal width • 3 classes • Setosa • Versicolor • Virginica
  • 45. SVM
  • 47.
  • 48.