Hey Guys, I think this Kaggle-competition is very useful for Scikit-Learn Beginners. I found that Random Forest Classifier works great with this model, but to better train the model, it is wise to use some clustering algorithm like Gaussian Mixture. I have uploaded my code. Please have a look and I am sure that you can learn a lot from it. Here's a link to this Competition https://www.kaggle.com/c/data-science-london-scikit-learn