Text Categorization on RCV1 (test)
0.0797Error Rateoh-CNN
Evaluation Results
| Method | Links | |
|---|---|---|
| oh-CNNUnlabeled data usage=1x200-dim CNN tv-embed.2016.02 | 0.0797 | |
| oh-2LSTMpUnlabeled data usage=2x100-dim LSTM tv-embed.2016.02 | 0.0924 | |
| wv-2LSTMpUnlabeled data usage=200-dim word2vec scaled2016.02 | 0.1018 | |
| wv-2LSTMpUnlabeled data usage=300-dim Google News word2vec2016.02 | 0.1062 | |
| wv-LSTMUnlabeled data usage=Pre-training2016.02 | 0.1465 | |
| L_MIXEDtraining=semi-supervised, objective=mixed objective function2020.09 | 6.23 | |
| LSTMtraining=semi-supervised2020.09 | 6.68 | |
| Adversarial + Virtual AdversarialBackbone=bidirectional LSTM2016.05 | 6.68 | |
| Virtual AdversarialBackbone=bidirectional LSTM2016.05 | 6.71 | |
| Adversarial + Virtual Adversarial2016.05 | 6.97 | |
| Virtual Adversarial2016.05 | 7.05 | |
| Adversarial2016.05 | 7.12 | |
| oh-LSTMtraining=semi-supervised2020.09 | 7.15 | |
| One-hot CNN2016.05 | 7.15 | |
| SA-LSTMtraining=semi-supervised2020.09 | 7.4 | |
| Baseline2016.05 | 7.4 | |
| One-hot CNNpretrained embeddings=CNN2016.05 | 7.71 | |
| L_MLTraining=Supervised2020.09 | 7.78 | |
| One-hot bi-LSTMpretrained embeddings=CNN and bidirectional LSTM2016.05 | 8.52 | |
| oh-CNNTraining=Supervised2020.09 | 9.17 | |
| CNNTraining=Supervised2020.09 | 10.44 | |
| Linear Model (TFIDF + SVM)Training=Supervised2020.09 | 10.68 | |
| Transductive SVM2016.05 | 10.77 | |
| NBLMdescription=Naive Bayes logisitic regression model2016.05 | 13.97 |