Research Article
A Study on Stock Graph Recognition Based on Wavelet Denoising and DTW Algorithm
Table 6
Comparison of various methods for optimal DTW distance analysis.
| ā | Zero-step ahead | One-step ahead | Two-step ahead | Preprocessing without wavelet denoising | Preprocessing with wavelet denoising | Preprocessing with MACD |
| Target graph I | 2.236 (0.3090) | 5.02 (0.1661) | 0.1197 (0.8930) | Target graph II | 2.637 (0.2749) | 0.7637 (0.5669) | 0.1682 (0.8560) | Target graph III | 3.592 (0.2177) | 0.9091 (0.5238) | 0.1171 (0.8951) | Target graph IV | 1.946 (0.3394) | 0.6371 (0.6108) | 0.1172 (0.8950) | Target graph V | 4.897 (0.1695) | 0.4918 (0.6703) | 0.076 (0.9293) |
|
|