Online Clustering of Bandits

Claudio Gentile, Shuai Li, Giovanni Zappella

2014-01-31preprintarXiv (Cornell University)164 citationsgreen

No readable copy stored yet — this record is metadata only.

Advanced Bandit Algorithms Research Recommender Systems and Techniques Reinforcement Learning in Robotics

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.

Citations in this library

Stored only when both papers are in the corpus — 164 citations exist upstream.

References

2002Finite-time Analysis of the Multiarmed Bandit Problem

2003Using confidence bounds for exploitation-exploration trade-offs

2011Improved Algorithms for Linear Stochastic Bandits

2011Contextual Bandits with Linear Payoff Functions

2009Exploration–exploitation tradeoff using variance estimates in multi-armed bandits

2011Unbiased offline evaluation of contextual-bandit-based news article recommendation algorithms

2007The Epoch-Greedy algorithm for contextual multi-armed bandits

2011Contextual Gaussian Process Bandit Optimization

Cited by

2016Distributed Clustering of Linear Bandits in Peer to Peer Networks

2017Factorization Bandits for Interactive Recommendation

2016Contextual Bandits in a Collaborative Environment

2017Bernoulli Rank-1 Bandits for Click Feedback

2018Learning Contextual Bandits in a Non-stationary Environment

2020Fast distributed bandits for online recommendation systems

2014Dynamic Clustering of Contextual Multi-Armed Bandits

2022Context‐based dynamic pricing with online clustering