This is the solution to the problem of the first round of ZS Data Science challenge 2019
I used this approach and secured 133rd rank in ZS Data Science Challenge 2019 round1. This model gave accuracy of 0.898 on Leaderboard.
NNFinal notebook is divided into two sections.
The intial part is data analysis where I have explored the data and gained insights on how the fill the missing values and what new features to generate.
This part focussed on imputing the missing values according to the inferences obtained in part1 and feature engineering. I used feed forward neural network for building the final model.
The predictions of the model can be found in NN_final.csv