Phishing has been a huge problem since the evolution of the Internet. It is kind of an Internet scam. Therefore, no antivirus or any kind of technical protection can completely eradicate this. But there has been research going on and conducted from intellects across the world to fight this online scam. Researches are coming up with different approaches to deal with this problem. The two main focused approaches taken by researchers to tackle phishing are Black Listing and Machine Learning. Machine Learning is a new and innovative way to tackle phishing. For this thesis I went with Machine Learning and heuristic based approach to tackle phishing. This thesis consist of a comparative study of different Machine Learning algorithm like Logistic Regression and ensemble algorithms like Adaboost and Gradientboost in order to know if ensemble algorithms can do a better prediction rather than standard machine Learning algorithms. The result obtained by ensemble algorithms were good but not as promising as expected.