Data Science Hackathon Project for Classification Problem
The aim of this project was to build a classifier to predict if a financial agent will get business for the company.
As part of the EDA, it was found that a lot of the data was missing. This made data imputation a challenge.
Also, there was a lot of scope for feature engineering.
My final solution used a label encoder and XGBoost to get a score of 65.28%. I am currently leading the leaderboard on 1st position.