This repository contains code of all the competitions I participate
#1. Customer Segmentation: I tried using feature engineering for this problem, however, no amount of adding features helped me increase my accuracy. Hence, I focussed on fine tuning my model. Model fine tuning got my rank from 23 to 8! I finised 12th in private leaderboard
#2. Healthcare Analytics II: We participated in a team for this competition. My partner(https://atif-hassan.github.io/) used neural networks and generated new features. I generated oof probabilities using 10-fold CV on lgbm. Then these two probabilities from nn and lgbm were blended using certain weightage given to each. This technique got us to rank 5 in public leaderboard and rank 9 in provate leaderboard.