Evolutionary Spiking Neural Networks for Solving Supervised Classification Problems

Table 4

Accuracy of design and testing on every configuration for Glass, Ionosphere, Iris Plant, Liver, Parkinson, and Wine datasets.

Dataset	Configuration	Design accuracy	Test accuracy
Glass	α1	0.2549 ± 0.1345	0.2404 ± 0.1433
	α2	0.6035 ± 0.0448	0.5374 ± 0.0688
	β1	0.4288 ± 0.0673	0.4002 ± 0.0590
	β2	0.6641 ± 0.0255	0.5947 ± 0.0436
	γ1	0.4895 ± 0.0574	0.4351 ± 0.0476
	γ2	0.7126 ± 0.0190	0.6186 ± 0.0413
Ionosphere	α1	0.8549 ± 0.0374	0.8374 ± 0.0295
	α2	0.9158 ± 0.0217	0.8669 ± 0.0267
	β1	0.8543 ± 0.0537	0.8137 ± 0.0708
	β2	0.9190 ± 0.0284	0.8724 ± 0.0241
	γ1	0.9351 ± 0.0182	0.8907 ± 0.0240
	γ2	0.9616 ± 0.0113	0.9015 ± 0.0201
Iris Plant	α1	0.8857 ± 0.1111	0.8663 ± 0.1269
	α2	0.9653 ± 0.0161	0.9386 ± 0.0210
	β1	0.9733 ± 0.0157	0.9382 ± 0.0217
	β2	0.9859 ± 0.0109	0.9362 ± 0.0163
	γ1	0.9794 ± 0.0123	0.9325 ± 0.0164
	γ2	0.9923 ± 0.0074	0.9358 ± 0.0261
Liver	α1	0.6820 ± 0.0406	0.6462 ± 0.0536
	α2	0.7352 ± 0.0245	0.6660 ± 0.0356
	β1	0.6834 ± 0.0481	0.6304 ± 0.0476
	β2	0.7461 ± 0.0224	0.6632 ± 0.0394
	γ1	0.7472 ± 0.0183	0.6723 ± 0.0302
	γ2	0.7636 ± 0.0196	0.6612 ± 0.0295
Parkinson	α1	0.8719 ± 0.0395	0.8281 ± 0.0568
	α2	0.9080 ± 0.0159	0.8596 ± 0.0285
	β1	0.8563 ± 0.0264	0.8033 ± 0.0519
	β2	0.8953 ± 0.0205	0.8503 ± 0.0353
	γ1	0.9025 ± 0.0266	0.8380 ± 0.0387
	γ2	0.9200 ± 0.0172	0.8494 ± 0.0377
Wine	α1	0.6881 ± 0.1549	0.6415 ± 0.1551
	α2	0.9063 ± 0.0441	0.8384 ± 0.0606
	β1	0.7375 ± 0.1126	0.6816 ± 0.1098
	β2	0.8895 ± 0.0641	0.7855 ± 0.0686
	γ1	0.9318 ± 0.0285	0.8620 ± 0.0491
	γ2	0.9638 ± 0.0164	0.8684 ± 0.0458