The effect of feature extraction and data sampling on credit card fraud detection

Journal of Big Data

Table 9 Performance scores for SMOTE Tomek

Learner	Step 1	Step 2	F1	AUC
Random Forest	PCA	SmoteTomek	0.818	0.913
XGBoost	PCA	SmoteTomek	0.729	0.933
LightGBM	PCA	SmoteTomek	0.445	0.937
CatBoost	PCA	SmoteTomek	0.700	0.928
Random Forest	SmoteTomek	PCA	0.807	0.918
XGBoost	SmoteTomek	PCA	0.609	0.938
LightGBM	SmoteTomek	PCA	0.368	0.931
CatBoost	SmoteTomek	PCA	0.593	0.927
Random Forest	SmoteTomek	CAE	0.899	0.970
XGBoost	SmoteTomek	CAE	0.852	0.970
LightGBM	SmoteTomek	CAE	0.478	0.967
CatBoost	SmoteTomek	CAE	0.741	0.940
Random Forest	CAE	SmoteTomek	0.855	0.922
XGBoost	CAE	SmoteTomek	0.702	0.929
LightGBM	CAE	SmoteTomek	0.509	0.925
CatBoost	CAE	SmoteTomek	0.606	0.947
Random Forest	SmoteTomek	None	0.872	0.918
XGBoost	SmoteTomek	None	0.813	0.923
LightGBM	SmoteTomek	None	0.693	0.938
CatBoost	SmoteTomek	None	0.746	0.928