Online Clustering of Bandits

Abstract

We introduce a novel algorithmic approach to content recommendation based on adaptive clustering of exploration-exploitation ("bandit") strategies. We provide a sharp regret analysis of this algorithm in a standard stochastic noise setting, demonstrate its scalability properties, and prove its effectiveness on a number of artificial and real-world datasets. Our experiments show a significant increase in prediction performance over state-of-the-art methods for bandit problems.

In E. Xing and T. Jebara (Eds.), Proceedings of 31st International Conference on Machine Learning, Journal of Machine Learning Research Workshop and Conference Proceedings, Vol.32 (JMLR W&CP-32), Beijing, China, Jun. 21-26, 2014 (ICML 2014), Submitted by Shuai Li (https://sites.google.com/site/shuailidotsli)

References in corpus (2)

• Spectral bandits for smooth graph functions
• Hierarchical Exploration for Accelerating Contextual Bandits

Cited by in corpus (40)

• Estimation-Action-Reflection: Towards Deep Interaction Between Conversational and Recommender Systems
• Distributed Clustering of Linear Bandits in Peer to Peer Networks
• Online Interactive Collaborative Filtering Using Multi-Armed Bandit with Dependent Arms
• Learning Contextual Bandits in a Non-stationary Environment
• Dynamically Expandable Graph Convolution for Streaming Recommendation
• Federated Linear Contextual Bandits
• Revealing graph bandits for maximizing local influence
• Simple Regret Minimization for Contextual Bandits
• Bilinear Bandits with Low-rank Structure
• Latent Bandits Revisited
• Graph Clustering Bandits for Recommendation
• Graph Neural Bandits
• RecSim NG: Toward Principled Uncertainty Modeling for Recommender Ecosystems
• Kernel Methods for Cooperative Multi-Agent Contextual Bandits
• Metadata-based Multi-Task Bandits with Bayesian Hierarchical Models
• Improved Algorithm on Online Clustering of Bandits
• Bandit Algorithms for Precision Medicine
• Cooperative Multi-Agent Bandits with Heavy Tails
• Meta-learning with Stochastic Linear Bandits
• Meta-Thompson Sampling
• Neural Bandit with Arm Group Graph
• Multi-facet Contextual Bandits: A Neural Network Perspective
• Regret Guarantees for Item-Item Collaborative Filtering
• Dynamic Global Sensitivity for Differentially Private Contextual Bandits
• Hierarchical Bayesian Bandits
• Local Clustering in Contextual Multi-Armed Bandits
• DeepNC: Deep Generative Network Completion
• Show Me the Whole World: Towards Entire Item Space Exploration for Interactive Personalized Recommendations
• Regret in Online Recommendation Systems
• Multitask Bandit Learning Through Heterogeneous Feedback Aggregation
• An Arm-Wise Randomization Approach to Combinatorial Linear Semi-Bandits
• Learning Networked Exponential Families with Network Lasso
• When and Whom to Collaborate with in a Changing Environment: A Collaborative Dynamic Bandit Solution
• Bandit algorithms for real-time data capture on large social medias
• No Regrets for Learning the Prior in Bandits
• Optimal Strategies for Graph-Structured Bandits
• Fast Distributed Bandits for Online Recommendation Systems
• The Use of Bandit Algorithms in Intelligent Interactive Recommender Systems
• Recommending with Recommendations
• Unifying Clustered and Non-stationary Bandits