Presentation is loading. Please wait.

Presentation is loading. Please wait.

[slides prises du cours cs294-10 UC Berkeley (2006 / 2009)]

Similar presentations


Presentation on theme: "[slides prises du cours cs294-10 UC Berkeley (2006 / 2009)]"— Presentation transcript:

1 [slides prises du cours cs UC Berkeley (2006 / 2009)]

2 Classification (reminder) X ! Y Anything: continuous ( ,  d, …) discrete ({0,1}, {1,…k}, …) structured (tree, string, …) … discrete: – {0,1}binary – {1,…k}multi-class – tree, etc.structured

3 Classification (reminder) X Anything: continuous ( ,  d, …) discrete ({0,1}, {1,…k}, …) structured (tree, string, …) …

4 Classification (reminder) X Anything: continuous ( ,  d, …) discrete ({0,1}, {1,…k}, …) structured (tree, string, …) … Perceptron Logistic Regression Support Vector Machine Decision Tree Random Forest Kernel trick

5 Regression X ! Y continuous: – ,  d Anything: continuous ( ,  d, …) discrete ({0,1}, {1,…k}, …) structured (tree, string, …) … 1

6

7

8

9

10

11

12

13

14

15

16

17

18

19 degree 15 overfitting!

20  Between two models / hypotheses which explain as well the data, choose the simplest one  In Machine Learning: ◦ we usually need to tradeoff between  training error  model complexity ◦ can be formalized precisely in statistics (bias- variance tradeoff, etc.)

21 training errormodel complexity

22

23  Logiciels: ◦ Weka (Java): ◦ RapidMiner (nicer GUI?): ◦ SciKit Learn (Python):  Livres: ◦ Pattern Classification (Duda, Hart & Stork) ◦ Pattern Recognition and Machine Learning (Bishop) ◦ Data Mining (Witten, Frank & Hall) ◦ The Elements of Statistical Learning (Hastie, Tibshirani, Friedman)  Programmer en python: ◦ cours cs188 de Dan Klein à Berkeley:

24

25 Kernel Regression Kernel regression (sigma=1)


Download ppt "[slides prises du cours cs294-10 UC Berkeley (2006 / 2009)]"

Similar presentations


Ads by Google