KaggleX Final Project Presenation with Parisa Zareapour | Kaggle

About the project: This project is a binary classification with the goal to predict whether the customers have defaulted. I trained and evaluated balanced_accuracy using XGBoost, LightGBM and HistGradientBoostClassifier. For the categorical features, I used ordinal and on-hot encoding and compared the performance. I also modeled the time-series portion of the data using a neural network.

About Parisa Zareapour: Graduated from my PhD in Physics in 2015. Has been working in finance as a quantitative analyst. In her job, she …

binary categorical classification customers encoding features hot kaggle lightgbm ordinal performance project series xgboost