Table 2 Experimental configurations of AeKNN
Model | Number of hidden layers | Number of neurons per layer | Architecture \({l_i}^n\) | ||||
|---|---|---|---|---|---|---|---|
Layer 1 | Layer 2 | Latent layer | Layer 4 | Layer 5 | |||
AeKNN1 | 1 | – | – | 32 | – | – | (32) |
AeKNN2 | 1 | – | – | 16 | – | – | (16) |
AeKNN3 | 1 | – | – | 8 | – | – | (8) |
AeKNN4 | 3 | 32 | – | 16 | – | 32 | (32,16,32) |
AeKNN5 | 5 | 32 | 16 | 8 | 16 | 32 | (32,16,8,16,32) |