Extreme Multi-label Classification on Wiki10-31K legacy (test)
89.74P@1CascadeXML
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CascadeXMLfeatures=DNN, ensemble_size=3 models, input_sequence_length=2562022.10 | 89.74 | 80.13 | 70.75 | |
| LightXMLfeatures=DNN, ensemble_size=3 models, input_sequence_length=2562022.10 | 89.67 | 79.06 | 69.87 | |
| CascadeXMLfeatures=DNN + tf-idf, ensemble_size=3 models, input_sequence_length=2562022.10 | 89.18 | 79.71 | 71.19 | |
| XR-Transformerfeatures=DNN + tf-idf, ensemble_size=3 models, input_sequence_length=2562022.10 | 88.69 | 80.17 | 70.91 | |
| X-Transformerfeatures=DNN + tf-idf, ensemble_size=9 models, input_sequence_length=2562022.10 | 88.26 | 78.51 | 69.68 | |
| AttentionXMLfeatures=DNN, ensemble_size=3 models, input_sequence_length=2562022.10 | 87.34 | 78.18 | 69.07 | |
| XR-Linear2022.10 | 85.75 | 75.79 | 66.69 | |
| Bonsai2022.10 | 84.52 | 73.76 | 64.69 | |
| Parabel2022.10 | 84.19 | 72.46 | 63.37 | |
| DISMEC2022.10 | 84.13 | 74.72 | 65.94 | |
| eXtremeText2022.10 | 83.66 | 73.28 | 64.51 |