Table 8.

The pruning results of LeNet-5 under different parameters.

$\mathit{\lambda }$	Fitness	Accuracy	Sparsity
0.1	0.02382	0.9901	0.8608
0.2	0.02978	0.9896	0.9031
0.4	0.04582	0.9871	0.9177
0.6	0.05516	0.9857	0.9319
0.8	0.06866	0.9843	0.9338
1	0.0790	0.9837	0.9373
2	0.1451	0.9791	0.9379
4	0.2518	0.9726	0.9439
6	0.2516	0.9686	0.9633
8	0.2610	0.9678	0.9714
10	0.2822	0.9648	0.9753