Toward Memory-Efficient and Interpretable Factorization Machines via Data and Model Binarization

Computer Science

• binarization 100%
• Feature Interaction 100%
• Experimental Result 50%
• Performance Evaluation 50%
• Gradient Descent 50%
• Interpretability 50%
• Feature Mapping 50%
• Feature Individual 50%
• Classification Rule 50%
• Binary Constraint 50%

Keyphrases

• Binarization 100%
• Memory Efficiency 100%
• Factorization Machines 100%
• Memory Cost 30%
• Feature Interaction 20%
• Performance Evaluation 10%
• Linear Time 10%
• Model Features 10%
• Adaptive Data 10%
• Binary Values 10%
• Nonlinear Features 10%
• Classification Datasets 10%
• Classification Task 10%
• Recent Approaches 10%
• Feature Mapping 10%
• Classification via Regression 10%
• Data Binning 10%
• Regression Task 10%
• Ranking Task 10%
• Adaptive Gradient Methods 10%
• Classification Rules 10%
• Equal Width Binning 10%
• Straight-through Estimator 10%
• Binning Method 10%
• Binary Constraints 10%
• One-bit 10%