Leveraging augmentation techniques for tasks with unbalancedness within the financial domain: a two-level ensemble approach

EPJ Data Science

Table 19 Classification results for the credit approval dataset

Evaluation metric	Naive Bayes	SGD	KNN	MLP	Random forest	SVM
Without augmentation
F1 score macro	0.7703	0.657	0.6283	0.7498	0.7824	0.5326
F1 score micro	0.8087	0.6811	0.6541	0.778	0.8088	0.5658
Precision minority class	0.786	0.7574	0.6765	0.7762	0.8164	0.5909
Recall minority class	0.8486	0.6943	0.7114	0.8343	0.8171	0.62
DL approach
F1 score macro	0.7138	0.629	0.7797	0.8884	0.9771	0.3674
F1 score micro	0.773	0.687	0.8084	0.8932	0.9771	0.4712
Precision minority class	0.8183	0.8017	0.8068	0.8759	0.9853	0.639
Recall minority class	0.7057	0.5629	0.8229	0.9343	0.9686	0.3629