Active learning using smooth relative regret approximations with applications

Keyphrases

• Active Learning 100%
• Relative Regret 100%
• Disagreement Coefficient 66%
• Learning Rate 33%
• Learning Algorithm 16%
• Learning Problems 16%
• Excess Risk 16%
• Difference Function 16%
• Pairwise Clustering 16%
• Absolute Error 16%
• Pairwise Preference 16%
• Label Complexity 16%
• Pool-based Active Learning 16%
• Passive Learning 16%
• Learning to Rank 16%
• Coefficient Bounds 16%
• Concept Classes 16%
• Learning Bound 16%
• Semi-supervised Clustering 16%
• Estimated Losses 16%
• Disagreement Measures 16%

Mathematics

• Constant Factor 100%
• Clustering 100%
• Absolute Error 100%
• Excess Risk 100%

Computer Science

• Approximation (Algorithm) 100%
• Active Learning 100%
• Learning Rate 33%
• Learning Problem 16%
• Pairwise Preference 16%
• Based Active Learning 16%
• Side Information 16%
• Constant Factor 16%
• Learning Algorithm 16%