A detailed comparison of performance scores achieved by Machine Learning and Deep Learning algorithms on 3 different Phishing datasets. 3 different feature selection and 2 different dimensionality reduction techniques are used for comparison. More than 97% accuracy achieved using the proposed technique.