High-Performance Machine Learning for Large-Scale Data Classification considering Class Imbalance

Table 9

The statistical indices of evaluating the epoch and accuracy using the Wine dataset for 50 times experiments.

Algorithm	Epoch		Accuracy
	Mean	Variance	Mean (%)	Variance
Improved BPNN	723.26	6266.69	94.4	28.31
Traditional BPNN	536.8	5974.37	67.3	372.87
Self-adaptive BPNN	1531	31924.12	91.4	55.2