Академический Документы
Профессиональный Документы
Культура Документы
● Find a linear h yperplane (decision b oundary) that will separate the d ata
02/03/2018 Introduction to Data Mining 2
Support Vector Machines
B1
B2
B2
B1
B2
B1
B2
b21
b22
margin
b11
b12
B1
! !
w• x + b = 0
! ! ! !
w • x + b = −1 w • x + b = +1
b11
! ! b12
! ⎧ 1 if w • x + b ≥ 1 2
f ( x ) = ⎨ ! ! Margin = !
⎩− 1 if w • x + b ≤ −1 || w ||
02/03/2018 Introduction to Data Mining 8
Linear SVM
● Linear m odel:
! !
! ⎧ 1 if w • x + b ≥ 1
f ( x ) = ⎨ ! !
⎩− 1 if w • x + b ≤ −1
Support vectors
x1 x2 y λ
0.3858 0.4687 1 65.5261
0.4871 0.611 -1 65.5261
0.9218 0.4103 -1 0
0.7382 0.8936 -1 0
0.1763 0.0579 1 0
0.4057 0.3529 1 0
0.9355 0.8132 -1 0
0.2146 0.0099 1 0
B1
B2
b21
b22
margin
b11
b12
Decision boundary:
! !
w • Φ( x ) + b = 0
02/03/2018 Introduction to Data Mining 17
● Optimization problem:
● Issues:
– What type of mapping function Φ should be
used?
– How to do the computation in high
dimensional space?
u Most computations involve dot product Φ(xi)• Φ(xj)
u Curse of dimensionality?
● Kernel T rick:
– Φ(xi)• Φ(xj) = K(xi, xj)
– K(xi, xj) is a kernel function (expressed in
terms of the coordinates in the original space)
u Examples: