Table 2.

Performance comparison between NRTPredictor and the other algorithms (Test dataset)

Method	Feature selection	No. of features	Accuracy %	Precision %	Recall %	F1-measure %
Lightgbm	F-score	150	96.53	94.27	93.37	93.78
XGBoost	F-score	430	97.88	96.48	96.33	96.39
SVM	F-score	180	96.34	94.02	93.35	93.66
RFC	F-score	210	94.41	93.09	85.51	88.14
Lightgbm	CV2	20,000	97.11	95.48	94.49	94.97
XGBoost	CV2	20,000	97.59	96.27	95.56	95.89
SVM	CV2	7000	94.22	93.45	89.92	91.27
RFC	CV2	14,000	89.71	88.02	85.43	85.37
Lightgbm	MIC	100	95.49	94.49	93.70	94.02
XGBoost	MIC	100	96.59	94.80	94.56	94.61
SVM	MIC	110	96.72	95.15	94.84	94.92
RFC	MIC	120	93.55	91.18	81.17	85.18
NRTPredictor	MIC	110	98.01	95.63	95.45	95.95