Emotion Classification on PRDECT-ID (test)
53.99AccuracyTextCNN
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| TextCNNevaluation_split=10% test split, sample_count=539 reviews2026.04 | 53.99 | 50.77 | 53.88 | 84.58 | |
| BiLSTM Baselineevaluation_split=10% test split, sample_count=539 reviews2026.04 | 51.21 | 43.54 | 48.29 | 82.12 | |
| BiLSTM Largeevaluation_split=10% test split, sample_count=539 reviews2026.04 | 51.21 | 42.4 | 48.28 | 82.39 | |
| BiLSTM Improvedevaluation_split=10% test split, sample_count=539 reviews2026.04 | 50.83 | 46.58 | 49.99 | 82.57 | |
| TF-IDF + Best AutoMLevaluation_split=20% held-out split, sample_count=1079 reviews2026.04 | 28.73 | 24.51 | 28.22 | 52.91 |