Feedforward Neural Network Cost Function

nnCostFunction_Feedforward.m

...
% Feedforward the neural network and return the cost in the variable J
summary = 0;
for j = 1:m

    y_label = y(j,:);

    a_one = X(j,:);
    a_one = [1 a_one];
    z_one = a_one * Theta1';
    a_two = sigmoid(z_one);
    m_temp = size(a_two, 1);
    a_two = [ones(m_temp, 1) a_two];
    z_two = a_two * Theta2';
    a_three = sigmoid(z_two);
    h = a_three;

    sum_temp = 0;
    sum_temp = (-y_label .* log(h)) - ((1 - y_label) .* log(1 - h));
    summary = summary + sum(sum_temp); 
...
end
...
% Add Regularization
regularization = (lambda/(2*m)) * (sum(sum(Theta1(:,2:end) .^ 2, 2)) + sum(sum(Theta2(:,2:end) .^ 2, 2)));
J = (summary/m) + regularization;
...