An Optimized Classification Algorithm by Neural Network Ensemble Based on PLS and OLS

Table 4

Comparison of every classifier’s accuracy for Test 2.

Network model	Weights	Training steps	1	2	3	4	5	SSE
Actual value	—	—	0	1	0	1	0	—
Subclassifier I  (BP NN)	0.3127	1205	−0.2163	1.5213	0.2027	0.6994	−0.1050	0.4610
Subclassifier II  (EBF NN)	0.2208	45	0.3255	1.4126	−0.1685	1.5012	0.2167	0.6027
Subclassifier III  (Elman NN)	0.4665	877	−0.1294	0.8927	−0.5101	1.2093	−0.0972	0.3417
PLS-Elman	—	408	0.1273	0.5792	−0.2216	1.1749	0.1132	0.2858
Assembling system	—	—	−0.0561	1.2041	−0.2118	1.1143	−0.0303	0.1036