Research Article
Is Deep Learning for Image Recognition Applicable to Stock Market Prediction?
Table 5
Accuracy comparison for CNNs with different dropout probabilities during the test period.
| ā | Dropout probability | Hit ratio | Specificity | Sensitivity |
| CNN1 | 0 | 0.85 | 0.9593 | 0.6971 | 0.25 | 0.67 | 0.6904 | 0.6507 | 0.5 | 0.66 | 0.6611 | 0.6508 |
| CNN2 | 0 | 0.62 | 0.6679 | 0.5878 | 0.25 | 0.68 | 0.6992 | 0.6744 | 0.5 | 0.67 | 0.6825 | 0.6209 |
| CNN3 | 0 | 0.64 | 0.9559 | 0.2487 | 0.25 | 0.64 | 0.9091 | 0.3012 | 0.5 | 0.62 | 0.6151 | 0.6302 |
| CNN4 | 0 | 0.66 | 0.6548 | 0.6872 | 0.25 | 0.62 | 0.6040 | 0.6375 | 0.5 | 0.62 | 0.6114 | 0.6393 |
|
|