Classification K-nearest neighbor classification D istance functions - - PowerPoint PPT Presentation

classification k nearest neighbor classification d
SMART_READER_LITE
LIVE PREVIEW

Classification K-nearest neighbor classification D istance functions - - PowerPoint PPT Presentation

Classification K-nearest neighbor classification D istance functions Choice of k Choice of k Leave-one-out cross validation K-fold cross validation Classification Error = Average classification error on K folds Linear Classification Linear


slide-1
SLIDE 1

Classification

slide-2
SLIDE 2
slide-3
SLIDE 3

K-nearest neighbor classification

slide-4
SLIDE 4

Distance functions

slide-5
SLIDE 5

Choice of k

slide-6
SLIDE 6

Choice of k

slide-7
SLIDE 7
slide-8
SLIDE 8

Leave-one-out cross validation

slide-9
SLIDE 9

K-fold cross validation

Classification Error = Average classification error on K folds

slide-10
SLIDE 10

Linear Classification

slide-11
SLIDE 11

Linear separability

slide-12
SLIDE 12
  • Real world problems: there may not exist a hyperplane

that separates cleanly

  • Solution to this “inseparability” problem: map data to

higher dimensional space

  • Called the “feature space”, as opposed to the original “input

space”

  • Inseparable training set can be made separable with proper

choice of feature space

Inseparability

slide-13
SLIDE 13

Feature map

slide-14
SLIDE 14

Linear classifier

slide-15
SLIDE 15

Linear classifier

slide-16
SLIDE 16

Good and bad linear classifiers

slide-17
SLIDE 17

Support Vector Machine

Two popular implementations

slide-18
SLIDE 18

Margin

slide-19
SLIDE 19

Margin

slide-20
SLIDE 20

Linear Support Vector Machine

slide-21
SLIDE 21

Inseparable case

slide-22
SLIDE 22

Linear SVM