Table 3. Comparison between the simulation results of different methods, for Example 2–1.
| Method | NOR | NOP | Training RMSE | Test RMSE |
|
|---|---|---|---|---|---|
| 1 | FJWNN | 3 | 41±2 (mean±std) |
0.000023±0.000014 (mean±std) |
0.000026±0.000026 (mean±std) |
| 2 | FWNN [4] | 3 | 27 | 0.0197 | 0.0226 |
| 3 | FWNN [4] | 3 | 43 | 0.0187 | 0.0202 |
| 4 | PRWNN [48] | -- | 48 | -- | 0.0102 |
| 5 | Type-2 FWNN with FCM [44] | 4 | 33 | 0.0167 | 0.0187 |
| 6 | FWNN[11] | 2 | 30 | 0.0067 | 0.0163 |
NOR, Number of rules; NOP, Number of model parameters; RMSE, Root mean square error; FWNN, Fuzzy wavelet neural network; PRWNN, Pipeline recurrent wavelet neural network; FCM, Fuzzy C-means clustering; -- No information is mentioned in the reference.