Medical Report Classification on PIFIR (test)
92.9AccuracyTAGCOS
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| TAGCOSStrategy=TAGCOS, Selection Budget=8 samples, Transfer Learning Protocol=PIFIR to CHIFIR, Fine-tuning=true2026.04 | 92.9 | 95.4 | 91.2 | 100 | 82.7 | |
| UncertaintyStrategy=Uncertainty, Selection Budget=8 samples, Transfer Learning Protocol=PIFIR to CHIFIR, Fine-tuning=true2026.04 | 90.5 | 93.8 | 90.9 | 96.8 | 88 | |
| RADSStrategy=RADS, Selection Budget=8 samples, Transfer Learning Protocol=PIFIR to CHIFIR, Fine-tuning=true2026.04 | 88.1 | 92.1 | 90.6 | 93.5 | 86.5 | |
| RandomStrategy=Random, Selection Budget=8 samples, Transfer Learning Protocol=PIFIR to CHIFIR, Fine-tuning=true2026.04 | 75.2 | 79.1 | 91.6 | 72.9 | 83.7 | |
| DiversityStrategy=Diversity, Selection Budget=8 samples, Transfer Learning Protocol=PIFIR to CHIFIR, Fine-tuning=true2026.04 | 73.8 | 84.9 | 73.8 | 100 | 88.3 | |
| LM-DPPStrategy=LM-DPP, Selection Budget=8 samples, Transfer Learning Protocol=PIFIR to CHIFIR, Fine-tuning=true2026.04 | 73.8 | 84.9 | 73.8 | 100 | 81.5 | |
| BatchBALDStrategy=BatchBALD, Selection Budget=8 samples, Transfer Learning Protocol=PIFIR to CHIFIR, Fine-tuning=true2026.04 | 71.4 | 79.3 | 85.2 | 74.2 | 74.2 |