Table 2. Experimental data distribution of the LEAD1.0 dataset.

		Normal (Real)	Abnormal (Real+Synthesized)	Total
LEAD1.0	Training	75,912	71,914 (47,914+24,000)	147,826
	Testing	18,978	11,978 (11,978+0)	30,956
	Total	94,890	83,892 (59,892+24,000)	178,782