Sigmoid function in NN

X = [ones(m, ) X];
temp = X * Theta1';

t = size(temp, );
temp = [ones(t, ) temp];

h = temp * Theta2';
[max_num, p] = max(h, [], );

Without Sigmoid function, Training Set Accuracy: 69.620000

 X = [ones(m, ) X];
 temp = X * Theta1';
 temp = sigmoid(temp);

 t = size(temp, );
 temp = [ones(t, ) temp];

 h = temp * Theta2';
 h = sigmoid(h);
 [max_num, p] = max(h, [], );

With Sigmoid function, the Training Set Accuracy: 97.520000