1. Ch. 18 – Learning Supplemental slides for CSE 327 Prof. Jeff Heflin

2. Decision Tree Learning function D EC -T REE -L EARN (examples,attribs,parent_examples) returns a decision tree if examples is empty then return P LURALITY -V ALUE (parent_examples) else if all examples have the same classification then return the classification else if attribs is empty then return P LURALITY -V ALUE (examples) else A  argmax a  attribs I MPORTANCE ( A, examples) tree  a new decision tree with root test A for each value v k of A do exs  {e : e  examples and e.A = v k } subtree  D EC -T REE -L EARN (exs,attribs – A, examples) add a branch to tree with label (A = v k ) and subtree subtree return tree From Figure 18.5, p. 702

3. Decision Tree Data Set ExampleColorSizeShape Goal Predicate X1bluesmallsquareno X2greenlargetriangleno X3redlargecircleyes X4greensmallsquareno X5yellowsmallcircleno X6redsmallcircleyes X7bluelargetriangleno X8redsmallsquareno

4. Decision Tree Result Shape? Color? No Yes +: X3,X6 -: X1,X2,X4,X5,X7,X8 +: -: X1,X4,X8 +: -: X2,X7 +: X3,X6 -: X5 +: X3,X6 -: +: -: X5 +: -: square triangle circle red blue green yellow

5. A Neuron

6. Perceptron Learning function P ERCEPTRON -L EARNING (examples,network) returns a perceptron hypothesis inputs: examples, a set of examples with input x and output y network, a perceptron with weights W j and activation function g repeat for each example (x,y) in examples do Err  y – g(in) for each j in 0..n W j  W j +   Err  g’(in)  x j until some stopping criteria is satisfied return N EURAL -N ET -H YPOTHESIS (network)

7. NETTalk OE_Y_AR … … 26 output units one layer of 80 hidden units 7x29 input units /r/

8. ALVINN Input is 30x32 pixels = 960 values 1 input pixel 5 hidden units 30 output units Sharp right Straight ahead Sharp left

9. SVM Kernels Non-linear separator in 2 dimensions: Mapped to 3 dimensions