机器学习作业之 Multi-class Classification and Neural Networks（week4）

1Regularied Logistic Regression

h = sigmoid(X*theta);

J = sum((-y).*log(h) - (1-y).*log(1-h))./m +(lambda) * sum(theta(2:end).^2)/(2*m);

grad = X'*(h-y)/m;
temp = grad (1); 
grad = grad + (lambda/m)*theta;
grad(1) = temp;

2 One-vs-all classifier training

  for c = 1:num_labels
	initial_theta = zeros(n+1,1);
 	options = optimset('GradObj', 'on', 'MaxIter', 50);
	[theta]= fmincg (@(t)(lrCostFunction(t,X,(y == c),lambda)),initial_theta,options);
	all_theta(c,:) = theta';
   end

3 One-vs-all classifier prediction

 [k,i] = max(sigmoid(X * all_theta'),[],2);
 p = i;

4 Neural Network Prediction Function

a1 = [ones(m,1) X];
z2 = Theta1 *( a1');
a2 = sigmoid(z2);
a2 = [ones(1, m);a2];
z3 = Theta2 * a2;
a3 = sigmoid(z3);

output =a3';  
[c,i] = max(output, [], 2);  
p = i;