World Journal of Surgical Oncology

Table 6 Comparison of before and after processing of data imbalances

From: Peritoneal cytology predicting distant metastasis in uterine carcinosarcoma: machine learning model development and validation

GBM		ROC-AUC	Threshold	Accuracy	Sensitivity	Specificity	Precision	F1
No treatment	Train	0.906	0.189	0.814	0.89	0.796	0.517	0.654
No treatment	Test	0.874	0.189	0.805	0.842	0.796	0.503	0.63
Resampling	Train	0.927	0.529	0.859	0.875	0.844	0.848	0.861
Resampling	Test	0.872	0.529	0.825	0.768	0.839	0.54	0.634
Weighted	Train	0.911	0.525	0.829	0.88	0.816	0.54	0.669
Weighted	Test	0.871	0.525	0.81	0.818	0.808	0.511	0.629

Back to article page

ISSN: 1477-7819

Contact us

General enquiries: journalsubmissions@springernature.com