Contextual Linear Bandits under Noisy Features: Towards Bayesian Oracles
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2017
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910642457280512 |
|---|---|
| author | Kim, Jung-hun Yun, Se-Young Jeong, Minchan Nam, Jun Hyun Shin, Jinwoo Combes, Richard |
| author_facet | Kim, Jung-hun Yun, Se-Young Jeong, Minchan Nam, Jun Hyun Shin, Jinwoo Combes, Richard |
| contents | We study contextual linear bandit problems under feature uncertainty, where the features are noisy and have missing entries. To address the challenges posed by this noise, we analyze Bayesian oracles given the observed noisy features. Our Bayesian analysis reveals that the optimal hypothesis can significantly deviate from the underlying realizability function, depending on the noise characteristics. These deviations are highly non-intuitive and do not occur in classical noiseless setups. This implies that classical approaches cannot guarantee a non-trivial regret bound. Therefore, we propose an algorithm that aims to approximate the Bayesian oracle based on the observed information under this model, achieving $\tilde{O}(d\sqrt{T})$ regret bound when there is a large number of arms. We demonstrate the proposed algorithm using synthetic and real-world datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_1703_01347 |
| institution | arXiv |
| publishDate | 2017 |
| record_format | arxiv |
| spellingShingle | Contextual Linear Bandits under Noisy Features: Towards Bayesian Oracles Kim, Jung-hun Yun, Se-Young Jeong, Minchan Nam, Jun Hyun Shin, Jinwoo Combes, Richard Artificial Intelligence Machine Learning We study contextual linear bandit problems under feature uncertainty, where the features are noisy and have missing entries. To address the challenges posed by this noise, we analyze Bayesian oracles given the observed noisy features. Our Bayesian analysis reveals that the optimal hypothesis can significantly deviate from the underlying realizability function, depending on the noise characteristics. These deviations are highly non-intuitive and do not occur in classical noiseless setups. This implies that classical approaches cannot guarantee a non-trivial regret bound. Therefore, we propose an algorithm that aims to approximate the Bayesian oracle based on the observed information under this model, achieving $\tilde{O}(d\sqrt{T})$ regret bound when there is a large number of arms. We demonstrate the proposed algorithm using synthetic and real-world datasets. |
| title | Contextual Linear Bandits under Noisy Features: Towards Bayesian Oracles |
| topic | Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/1703.01347 |