13-PracticalMachineLearning

13-PracticalMachineLearning - Prac╩l Machine Learning...

Info iconThis preview shows pages 1–24. Sign up to view the full content.

View Full Document Right Arrow Icon
Prac%cal’Machine’Learning’ Verena’Kaynig-Fi4kau’([email protected])’ We’are’drowning’in’informa%on’and’ starving’for’knowledge’ John’ Naisbi4’
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Machine’Learning’ Analyze’training’data’ Make’predic%ons’for’new’unseen’data:’ supervised’learning’ Find’pa4erns:’ unsupervised’learning’
Background image of page 2
Machine’Learning’ Supervised’Learning’ SVM,’Decision’Tree,’Boos%ng,’Random’Forest’ Unsupervised’Learning’ K-means,’mean’shi±’
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Supervised’Learning’ x1’ x2’ data’points’ labels’ features’ separa%ng’ hyper’plane’
Background image of page 4
Features’are’important’ weight’ roundness’ ?
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Features’are’important’ color’ shape’
Background image of page 6
Google’s±Self-Driving±Car±
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Car’Features ’ Laser’scan’ Intensity’model’ Eleva%on’model’ Camera’vision’ 2D’sta%onary’map’ Lane’model’
Background image of page 8
So’just’measure’everything?’ More’features’=’be4er’classiFca%on?’ Prac%cal’issues:’ Data’volume,’computa%on’overhead’ Theore%cal’issues:’ Generaliza%on’performance’ Curse’of’dimensionality’
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Supervised’Learning’ x1’ x2’ data’points’ labels’ features’ separa%ng’ hyper’plane’
Background image of page 10
Perceptron’ x:’data’point’ y:’label’ w:’weight’vector’ b:’bias’ w’ -1 +1 x1’ x2’ x3’ -1’ b’ w3’ w2’ w1’
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
The’XOR’Problem’ x1’ x2’ x3’
Background image of page 12
Support’Vector’Machine’ Widely’used’for’all’sorts’of’classiFca%on’ problems’ www.clopinet.com/isabelle/Projects/SVM/applist.html Some’people’say’it’is’the’best’of’the’shelf’ classiFer’out’there’
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Maximum’Margin’Classifca%on’ x1’ x2’ x1’ x2’
Background image of page 14
What’about’outliers?’ x2’ x1’ :’slack’variables’
Background image of page 15

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
x=0’ XOR’problem’revised’ Did’we’add’informa%on’to’make’the’problem’ seperable?’
Background image of page 16
Polynomial’Kernel’in’3D’
Background image of page 17

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Quadra%c’Kernel’
Background image of page 18
Polynomial:’ Radial’basis’func%on’(RBF):’ Kernel’Func%ons’
Background image of page 19

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Kernel’Trick’for’SVMs’ Arbitrary’many’dimensions’ Li4le’computa%onal’cost’ Maximal’margin’helps’with’curse’of’ dimensionality’
Background image of page 20
SVM’Applet’ h4p://www.ml.inf.ethz.ch/educa%on/ lectures_and_seminars/annex_estat/ClassiFer/ JSupportVectorApplet.html
Background image of page 21

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Tips’and’Tricks’ SVMs’are’ not ’scale’invariant’ Check’if’your’library’normalizes’by’default’ Normalize’your’data’ mean:’0’,’ std:’1’ map’to’[0,1]’or’[-1,1]’ Normalize’test’set’in’same’way!’
Background image of page 22
Tips’and’Tricks’ RBF’kernel’is’a’good’default’ For’parameters’try’exponen%al’sequences’ Read:’ ’ Chih-Wei’Hsu’et’al.,’“ A’Prac%cal’Guide’to’ ’ Support’Vector’Classi±ca%on ”,’ ’ Bioinforma%cs’(2010)’
Background image of page 23

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 24
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

Page1 / 84

13-PracticalMachineLearning - Prac╩l Machine Learning...

This preview shows document pages 1 - 24. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online