Research Article
A Simple Method of Residential Electricity Load Forecasting by Improved Bayesian Neural Networks
Table 8
Results of time-series BNN under different input vectors.
| | TS-BNN (Hidden neurons 4) Input vectors | Training set (60%) | Test set (20%) | Computing time (s) | | MSE | R | MAPE (%) | MSE | R | MAPE (%) |
| | 9.97e-2 | 8.61e-1 | 11.67 | 1.02e-1 | 8.55e-1 | 11.78 | 15 |
| | 1.07e-1 | 8.55e-1 | 11.82 | 1.11e-1 | 8.58e-1 | 12.49 | 7 | | 1.08e-1 | 8.55e-1 | 11.60 | 1.05e-2 | 8.61e-1 | 11.45 | 9 | | 1.04e-1 | 8.60e-1 | 11.76 | 1.14e-1 | 8.52e-1 | 11.13 | 15 | | 1.01e-1 | 8.66e-1 | 11.91 | 1.09e-1 | 8.53e-1 | 11.44 | 5 |
|
|
