Text Classification on MR, CR, MPQA, SUBJ, TREC fine-tuning Noise Rate 40%
83.54MR AccuracyREGSL
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| REGSLArchitecture=three-layer feed forward network, Pre-trained=SST, Noise Rate=40%, Number of random seeds=52021.11 | 83.54 | 79.89 | 84.15 | 72.48 | 74.8 | |
| l2-PGMArchitecture=three-layer feed forward network, Pre-trained=SST, Noise Rate=40%, Number of random seeds=52021.11 | 83.45 | 79.47 | 84.03 | 72.14 | 73.48 | |
| Fine-tuningArchitecture=three-layer feed forward network, Pre-trained=SST, Noise Rate=40%, Number of random seeds=52021.11 | 82.91 | 77.67 | 82.85 | 90.78 | 70.8 |