Table 7.

Cross-validation performance of the proposed ALC and other classifiers on the MNIST dataset (mean over 10-folds).

Metric	ALC	XGB	SVM	MLP	LR
Loss	0.0000	0.0581	0.0076	0.0473	0.0137
Accuracy	0.9975	0.9421	0.9967	0.9900	0.9967
Precision	0.9970	0.9828	0.9953	0.9906	0.9953
Recall	0.9967	0.9802	0.9967	0.9900	0.9967
F1-Score	0.9987	0.9800	0.9967	0.9900	0.9967
Overfitting	0.0025%	0.0571%	0.0050%	0.0100%	0.0050%
Time (sec.)	6.18	2.35	5.38	5.22	5.61