Scalable, efficient, stepwise-optimal feature elimination in support vector machines

Computer Science

• Support Vector Machine 100%
• Feature Selection 100%
• Feature Space 100%
• Feature Extraction 100%
• Memory Requirement 33%
• Training Data 33%
• Generalization Performance 33%
• Classification Accuracy 33%
• Training Error 33%
• Non-Separable 33%
• Data Domain 33%

Keyphrases

• Weight Magnitude 66%
• Lightweight Scheme 33%
• Classification Complexity 33%
• Biomedical Context 33%
• Data Domain 33%
• Current Solution 33%
• Margin Characteristic 33%
• Classifier Generalization 33%