Text Classification on Amazon Review 1000 labels (test)
39.17Top-1 Error RateSoftMatch/FreeMatch
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SoftMatch/FreeMatchFramework=+SR2023.10 | 39.17 | — | — | |
| FlexMatchFramework=+SR2023.10 | 41.11 | — | — | |
| SoftMatch/FreeMatchFramework=Base2023.10 | 42.21 | — | — | |
| SimMatch# Labels=1000, Backbone=Bert-Base2023.05 | 42.21 | — | — | |
| SoftMatch# Labels=1000, Backbone=Bert-Base2023.05 | 42.21 | — | — | |
| FlexMatchFramework=Base2023.10 | 42.25 | — | — | |
| FlexMatch# Labels=1000, Backbone=Bert-Base2023.05 | 42.25 | — | — | |
| AdaMatch# Labels=1000, Backbone=Bert-Base2023.05 | 42.27 | — | — | |
| ILL# Labels=1000, Backbone=Bert-Base2023.05 | 42.32 | — | — | |
| FreeMatch# Labels=1000, Backbone=Bert-Base2023.05 | 42.64 | — | — | |
| FixMatch# Labels=1000, Backbone=Bert-Base2023.05 | 43.05 | — | — | |
| Dash# Labels=1000, Backbone=Bert-Base2023.05 | 43.09 | — | — | |
| CoMatch# Labels=1000, Backbone=Bert-Base2023.05 | 43.36 | — | — | |
| Pseudo LabelFramework=+SR2023.10 | 44.21 | — | — | |
| VAT# Labels=1000, Backbone=Bert-Base2023.05 | 46.54 | — | — | |
| Pseudo LabelFramework=Base2023.10 | 47 | — | — | |
| Pseudo-Label# Labels=1000, Backbone=Bert-Base2023.05 | 47 | — | — | |
| Mean-Teacher# Labels=1000, Backbone=Bert-Base2023.05 | 47.66 | — | — | |
| MixMatch# Labels=1000, Backbone=Bert-Base2023.05 | 61.69 | — | — | |
| SemiReward (Mean)Framework=+SR2023.10 | — | 2.32 | 2.92 |