Classification of lung cancer severity using gene expression data based on deep learning

Table 8 Accuracy evaluation of classical ML models using different training-testing methods with 5000 features

Dataset	Training – testing method	SVM	RF	KNN	AdaBoost
LUAD	10-fold cross-validation	0.68	0.56	0.56	0.58
	5-fold cross-validation	0.67	0.58	0.56	0.58
	80–20 splitting	0.79	0.65	0.62	0.66
	70–30 splitting	0.73	0.69	0.59	0.62
LUSC	10-fold cross-validation	0.698	0.594	0.559	0.58
	5-fold cross-validation	0.696	0.596	0.554	0.567
	80–20 splitting	0.74	0.65	0.62	0.59
	70–30 splitting	0.69	0.62	0.60	0.51

ISSN: 1472-6947