Research Article
A Hybrid Model Using PCA and BP Neural Network for Time Series Prediction in Chinese Stock Market with TOPSIS Analysis
Table 6
Parameter configuration for the BP neural network.
| | Number of neurons in hidden layer | Learning rate | Activation function | Experiment ID | Average | Experiment average |
| | 5 | 0.1 | Sigmoid | 1 | 1.6478 | 1.6143 | | 5 | 0.1 | Sigmoid | 2 | 1.8413 | | 5 | 0.1 | Sigmoid | 3 | 1.5479 | | 5 | 0.1 | Sigmoid | 4 | 1.4202 | | 5 | 0.1 | Sigmoid | 1 | 1.3254 | 1.4361 | | 5 | 0.1 | Sigmoid | 2 | 1.5648 | | 5 | 0.1 | Sigmoid | 3 | 1.7411 | | 5 | 0.1 | Sigmoid | 4 | 1.1131 | | 10 | 0.1 | Sigmoid | 1 | 1.1910 | 1.3285 | | 10 | 0.1 | Sigmoid | 2 | 1.5585 | | 10 | 0.1 | Sigmoid | 3 | 1.4521 | | 10 | 0.1 | Sigmoid | 4 | 1.1124 | | 15 | 0.1 | Sigmoid | 1 | 1.6425 | 1.5435 | | 15 | 0.1 | Sigmoid | 2 | 1.5788 | | 15 | 0.1 | Sigmoid | 3 | 1.4726 | | 15 | 0.1 | Sigmoid | 4 | 1.4801 |
|
|
Bold values signify the best results.
|
