A Practical Method for Solving Contextual Bandit Problems Using Decision Trees

TL;DR

This paper introduces a practical, decision tree-based method for solving contextual bandit problems that requires minimal domain expertise and employs a bootstrap approach for exploration-exploitation balancing.

Contribution

It presents a simple, interpretable algorithm using decision trees and bootstrap-based Thompson sampling for effective contextual bandit solutions with minimal tuning.

Findings

Performs well on multiple datasets

Requires little domain knowledge

Outperforms some existing methods

Abstract

Many efficient algorithms with strong theoretical guarantees have been proposed for the contextual multi-armed bandit problem. However, applying these algorithms in practice can be difficult because they require domain expertise to build appropriate features and to tune their parameters. We propose a new method for the contextual bandit problem that is simple, practical, and can be applied with little or no domain expertise. Our algorithm relies on decision trees to model the context-reward relationship. Decision trees are non-parametric, interpretable, and work well without hand-crafted features. To guide the exploration-exploitation trade-off, we use a bootstrapping approach which abstracts Thompson sampling to non-Bayesian settings. We also discuss several computational heuristics and demonstrate the performance of our method on several datasets.