# Context Attentive Bandits: Contextual Bandit with Restricted Context

## TL;DR

This paper introduces a new variant of the contextual bandit problem where only limited features are accessible at each step, proposing algorithms tailored for such restricted contexts and demonstrating their effectiveness on real datasets.

## Contribution

The paper formulates the contextual bandit with restricted context and develops two novel algorithms, TSRC and WTSRC, for stationary and nonstationary environments respectively.

## Key findings

- Proposed algorithms outperform baselines on real datasets.
- TSRC and WTSRC effectively handle limited feature access.
- Algorithms adapt to both stationary and nonstationary settings.

## Abstract

We consider a novel formulation of the multi-armed bandit model, which we call the contextual bandit with restricted context, where only a limited number of features can be accessed by the learner at every iteration. This novel formulation is motivated by different online problems arising in clinical trials, recommender systems and attention modeling. Herein, we adapt the standard multi-armed bandit algorithm known as Thompson Sampling to take advantage of our restricted context setting, and propose two novel algorithms, called the Thompson Sampling with Restricted Context(TSRC) and the Windows Thompson Sampling with Restricted Context(WTSRC), for handling stationary and nonstationary environments, respectively. Our empirical results demonstrate advantages of the proposed approaches on several real-life datasets