costFunctionJ.m

function J = costFunctionJ(X, y, theta)

% X is the "design matrix" containg our training examples
% y is the class labels

m = size(X,1);         % numbedr of training examples
predictions = X*theta; % predictions of hypothesis on all m examples
sqrErrors = (predictions - y) .^2; % squared errors

J= 1/(2*m) * sum(sqrErrors);