Online Spectral Learning on a Graph with Bandit Feedback

Keyphrases

• Information Settings 100%
• Spectral Learning 100%
• Bandit Feedback 100%
• Online Learning 75%
• Full Information 75%
• Label Information 75%
• Multi-class Classification 75%
• Oracle 50%
• Graph Laplacian 50%
• Class Label 50%
• Online Learning Algorithm 50%
• Regret Bounds 50%
• Spectral Information 25%
• Eigenvector 25%
• Function Learning 25%
• Binary Classification 25%
• Partial Information 25%
• Multi-class 25%
• Single-bit 25%
• Class number 25%
• Round numbers 25%
• Bandits 25%
• Graph Datasets 25%
• Partial Labels 25%
• Upper Confidence Bound 25%
• Benchmark Graphs 25%
• Node Classification 25%
• Bandit Algorithms 25%
• T-factor 25%

Computer Science

• Online Learning 100%
• Classification Algorithm 60%
• Multiclass Classification 60%
• Multi Class Classification 60%
• Learning Algorithm 40%
• Laplace Operator 40%
• Spectral Information 20%
• Binary Classification 20%
• Eigenvector 20%
• Application Scenario 20%
• Confidence Bound 20%
• Node Classification 20%