This file is for Humana Case Competition in 2022. In this competition, we are supposed to use provided features to identify those members facing home insecurity issues (Classification using Python).
The basic setting and data dictionary provided by the committee are contained in this file. One should be able to understand what we are trying to accomplish by reading these files.
In specific, one should open the "data cleaning" file first, and then apply the csv file of cleaned dataset from this file to build random forest model and XGBoost model in other files.
Finally, you can open the "after tuning hyperparameter" file to see how we implement the optimzed hyperparameters and extract final submissions.
The ranking is based on ROC curve and AUC metric, and we reached an AUC-ROC score of 0.75 (ranked 24th finally).