How to score 0.8134 in the Titanic Kaggle challenge

Hello everyone, I’ve been, like all of you, digging in the data and playing with it and I’ve recently achieved a score of 0.8134. I thought it’d be interesting if I shared my solution : So here is a guided tutorial in which I go step by step presenting each level of the pipeline. It goes from data analysis right up to hyper-parameters tuning. I’d love to have you feedback on it. So if you have also have any improvement to suggest (about the preprocessing, the modeling, etc…) do not hesitate! Looking forward to hearing from you! Cheers!


Link to Full Article: How to score 0.8134 in the Titanic Kaggle challenge