Explainable machine learning models for Medicare fraud detection

Journal of Big Data

Table 10 Mean AUC values by classifier and number of features for ten iterations of fivefold cross validation, for classifying the Medicare Part D data (Part 2)

Features classifier	15	20	25	30	82
CatBoost	0.9436	0.9560	0.9567	0.9588	0.9587
ET	0.8294	0.8323	0.8352	0.8429	0.8116
LightGBM	0.7505	0.7929	0.7913	0.8311	0.8455
Logistic Regression	0.9006	0.9143	0.9133	0.8620	0.8536
Random Forest	0.8315	0.8288	0.8316	0.8496	0.7909
XGBoost	0.9472	0.9390	0.9447	0.9449	0.9426