List your company
ResearchFormerly WizWand

Text Categorization on RCV1 (test)

0.0797Error Rateoh-CNN
-0.4759123.2744697.0248510.775231Feb 7, 2016Nov 12, 2016Aug 18, 2017May 24, 2018Feb 27, 2019Dec 3, 2019Sep 8, 2020
Updated 8mo ago

Evaluation results

MethodLinks
oh-CNNUnlabeled data usage=1x200-dim CNN tv-embed.2016.020.0797
oh-2LSTMpUnlabeled data usage=2x100-dim LSTM tv-embed.2016.020.0924
wv-2LSTMpUnlabeled data usage=200-dim word2vec scaled2016.020.1018
wv-2LSTMpUnlabeled data usage=300-dim Google News word2vec2016.020.1062
wv-LSTMUnlabeled data usage=Pre-training2016.020.1465
L_MIXEDtraining=semi-supervised, objective=mixed objective function2020.096.23
LSTMtraining=semi-supervised2020.096.68
Adversarial + Virtual AdversarialBackbone=bidirectional LSTM2016.056.68
Virtual AdversarialBackbone=bidirectional LSTM2016.056.71
Adversarial + Virtual Adversarial2016.056.97
Virtual Adversarial2016.057.05
Adversarial2016.057.12
oh-LSTMtraining=semi-supervised2020.097.15
One-hot CNN2016.057.15
SA-LSTMtraining=semi-supervised2020.097.4
Baseline2016.057.4
One-hot CNNpretrained embeddings=CNN2016.057.71
L_MLTraining=Supervised2020.097.78
One-hot bi-LSTMpretrained embeddings=CNN and bidirectional LSTM2016.058.52
oh-CNNTraining=Supervised2020.099.17
CNNTraining=Supervised2020.0910.44
Linear Model (TFIDF + SVM)Training=Supervised2020.0910.68
Transductive SVM2016.0510.77
NBLMdescription=Naive Bayes logisitic regression model2016.0513.97