Imbalanced Regression on STS-B-DIR Few-shot (test)
0.781MSEVIR
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| VIRBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 0.781 | 0.708 | 0.845 | 0.861 | |
| RANKSIMBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 0.804 | 0.712 | 0.827 | 0.849 | |
| LDS + FDSBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 0.942 | 0.755 | 0.766 | 0.749 | |
| LDSBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 0.955 | 0.772 | 0.762 | 0.743 | |
| INFER NOISEBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 0.967 | 0.741 | 0.756 | 0.753 | |
| VANILLABackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 0.984 | 0.771 | 0.752 | 0.75 | |
| FDSBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 1.012 | 0.756 | 0.742 | 0.748 | |
| DEEP ENSEMBLEBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 1.032 | 0.792 | 0.75 | 0.746 | |
| INVBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 1.046 | 0.78 | 0.732 | 0.714 | |
| DERBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 1.055 | 0.809 | 0.742 | 0.739 | |
| LDS + FDS + DERBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 1.086 | 0.842 | 0.731 | 0.699 | |
| VAEBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 1.102 | 0.767 | 0.749 | 0.743 | |
| SMOTERBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 1.154 | 0.861 | 0.706 | 0.691 | |
| SMOGNBackbone=BiLSTM, Word Embeddings=GloVe-300D, Weighting Strategy=SQINV, Batch Size=256, Optimizer=Adam2023.06 | 1.175 | 0.848 | 0.692 | 0.67 |