function [theta, J_history] = gradientDescentMulti(X, y, theta, alpha, num_iters)
%GRADIENTDESCENTMULTI Performs gradient descent to learn theta
%   theta = GRADIENTDESCENTMULTI(x, y, theta, alpha, num_iters) updates theta by
%   taking num_iters gradient steps with learning rate alpha

% Initialize some useful values
m = length(y); % number of training examples
J_history = zeros(num_iters, 1);

for iter = 1:num_iters

    % ====================== YOUR CODE HERE ======================
    % Instructions: Perform a single gradient step on the parameter vector
    %               theta.
    %
    % Hint: While debugging, it can be useful to print out the values
    %       of the cost function (computeCostMulti) and gradient here.
    %
    theta_old = theta;
    hipoteza = X * theta_old;
    diff = hipoteza - y;

    for j = 1:length(theta)
        predvsota(j) = diff(j) .* X(j);
        vsota = sum(predvsota);
        theta(j) = theta_old(j) - (alpha * (1 / m)) * vsota;
    end;

    % ============================================================

    % Save the cost J in every iteration
    J_history(iter) = computeCostMulti(X, y, theta);

end

end