MAKING PRE-TRAINED LANGUAGE MODELS GREAT ON TABULAR PREDICTION

Keyphrases

• Feature Value 100%
• Deep Neural Network 100%
• Tabular Data 100%
• Pre-trained Language Model 100%
• Data Prediction 66%
• Numerical Characteristics 66%
• Relative Magnitude 33%
• Image Processing 33%
• Comprehensive Experiment 33%
• Scalar 33%
• Language Processing 33%
• Transferability 33%
• Knowledge Transfer 33%
• Gradient Boosting Decision Tree 33%
• Text Representation 33%
• Language Model 33%
• Token 33%
• Decision Tree Model 33%
• Tokenization 33%
• Representation Space 33%
• Classification Task 33%
• Feature Attention 33%
• Prediction Task 33%
• Domain-specific Languages 33%
• Corresponding Features 33%
• Diverse Predictions 33%
• Regression Task 33%

Computer Science

• Deep Neural Network 100%
• Pre-Trained Language Models 100%
• Image Processing 33%
• Text Representation 33%
• Decision Tree Model 33%
• Classification Task 33%
• Language Modeling 33%
• Language Processing 33%
• Regression Task 33%
• Representation Space 33%
• Lexical Tokenization 33%

Engineering

• Deep Neural Network 100%
• Relative Magnitude 33%
• Classification Task 33%
• Model Tree 33%