Table 5 How does the number of selected features affect the prediction accuracy
From: Short-term stock market price trend prediction using a comprehensive deep learning system
Number of selected features | 5 principal components | 10 principal components | 15 principal components | 20 principal components | 29 selected features |
|---|---|---|---|---|---|
Accuracy | 89.03% | 89.35% | 89.39% | 89.30% | 90.29% |