Coursera_ML_Week3 Logistic Regression Cost and Gradient calculation with/without regularization of theta parameters h -> sigmoid function