predict.m

function p = predict(theta, X)
  %PREDICT Predict whether the label is 0 or 1 using learned logistic 
  %regression parameters theta
  %   p = PREDICT(theta, X) computes the predictions for X using a 
  %   threshold at 0.5 (i.e., if sigmoid(theta'*x) >= 0.5, predict 1)

  probability = sigmoid(X * theta);
  p = probability >= 0.5;
end