Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Nilsson, Hannes, Johansson, Rikard, Åkerblom, Niklas, Chehreghani, Morteza Haghir
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2402.06963
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915649574404096
author Nilsson, Hannes
Johansson, Rikard
Åkerblom, Niklas
Chehreghani, Morteza Haghir
author_facet Nilsson, Hannes
Johansson, Rikard
Åkerblom, Niklas
Chehreghani, Morteza Haghir
contents We propose a new framework for contextual multi-armed bandits based on tree ensembles. Our framework adapts two widely used bandit methods, Upper Confidence Bound and Thompson Sampling, for both standard and combinatorial settings. As part of this framework, we propose a novel method of estimating the uncertainty in tree ensemble predictions. We further demonstrate the effectiveness of our framework via several experimental studies, employing XGBoost and random forests, two popular tree ensemble methods. Compared to state-of-the-art methods based on decision trees and neural networks, our methods exhibit superior performance in terms of both regret minimization and computational runtime, when applied to benchmark datasets and the real-world application of navigation over road networks.
format Preprint
id arxiv_https___arxiv_org_abs_2402_06963
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Tree Ensembles for Contextual Bandits
Nilsson, Hannes
Johansson, Rikard
Åkerblom, Niklas
Chehreghani, Morteza Haghir
Machine Learning
Artificial Intelligence
We propose a new framework for contextual multi-armed bandits based on tree ensembles. Our framework adapts two widely used bandit methods, Upper Confidence Bound and Thompson Sampling, for both standard and combinatorial settings. As part of this framework, we propose a novel method of estimating the uncertainty in tree ensemble predictions. We further demonstrate the effectiveness of our framework via several experimental studies, employing XGBoost and random forests, two popular tree ensemble methods. Compared to state-of-the-art methods based on decision trees and neural networks, our methods exhibit superior performance in terms of both regret minimization and computational runtime, when applied to benchmark datasets and the real-world application of navigation over road networks.
title Tree Ensembles for Contextual Bandits
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2402.06963