Research Article
Time Series Prediction Based on Complex-Valued S-System Model
Table 8
The RMSE performances of our method with different training sets for SSEI and NASDAQ prediction problems.
| Datasets | | 500 | 1000 | 1500 | 2000 |
| SSEI | Average | 0.0069246 | 0.006561 | 0.006493 | 0.006469 | Best | 0.006786 | 0.006445 | 0.006418 | 0.006406 | Worse | 0.007541 | 0.006863 | 0.006568 | 0.006505 | Standard deviation | 0.000175 | 0.000106 | 3.76E − 05 | 3.24E − 05 | NASDAQ index | Average | 0.006493 | 0.006483 | 0.006098 | 0.006087 | Best | 0.006202 | 0.006201 | 0.005988 | 0.006032 | Worse | 0.006949 | 0.006812 | 0.006243 | 0.006149 | Standard deviation | 0.000198 | 0.000173 | 6.42E − 05 | 3.87E − 05 |
|
|