Table 4 Performance comparison for stage 1 cultural context recognition CNN ablation study.
From: A novel model for cultural-based classification of liberal arts using deep reinforcement learning
Configuration | Accuracy (%) | Precision | Recall | F-Measure |
|---|---|---|---|---|
Proposed (PReLU + Hybrid) | 96.9558 | 0.9667 | 0.9697 | 0.9681 |
ReLU + Hybrid | 95.3569 | 0.9498 | 0.9533 | 0.9515 |
PReLU + MaxPool | 96.8555 | 0.9661 | 0.9687 | 0.9673 |
PReLU + AvgPool | 95.6578 | 0.9527 | 0.9568 | 0.9547 |
