A Survey on Contextual Multi-armed Bandits

TL;DR

This survey reviews various stochastic and adversarial contextual bandit algorithms, analyzing their assumptions and regret bounds to provide a comprehensive overview of the field.

Contribution

It offers a systematic comparison of different algorithms, highlighting their theoretical guarantees and assumptions in the contextual bandit setting.

Findings

Different algorithms have varying regret bounds and assumptions.

The survey identifies key challenges and open problems in the field.

It provides a structured overview of the state-of-the-art methods.

Abstract

In this survey we cover a few stochastic and adversarial contextual bandit algorithms. We analyze each algorithm's assumption and regret bound.