Multi-facet Contextual Bandits: A Neural Network Perspective
This addresses a novel problem in recommender systems for domains like E-commerce and healthcare, but it appears incremental as it extends contextual bandits with neural networks.
The paper tackles the problem of multi-facet contextual bandits, where multiple bandits represent different user aspects, and aims to select arms from each to maximize reward, with applications in E-commerce and healthcare. It proposes the MuFasa algorithm, which achieves a near-optimal regret bound of Õ((K+1)√T) and outperforms baselines in experiments on real-world datasets.
Contextual multi-armed bandit has shown to be an effective tool in recommender systems. In this paper, we study a novel problem of multi-facet bandits involving a group of bandits, each characterizing the users' needs from one unique aspect. In each round, for the given user, we need to select one arm from each bandit, such that the combination of all arms maximizes the final reward. This problem can find immediate applications in E-commerce, healthcare, etc. To address this problem, we propose a novel algorithm, named MuFasa, which utilizes an assembled neural network to jointly learn the underlying reward functions of multiple bandits. It estimates an Upper Confidence Bound (UCB) linked with the expected reward to balance between exploitation and exploration. Under mild assumptions, we provide the regret analysis of MuFasa. It can achieve the near-optimal $\widetilde{ \mathcal{O}}((K+1)\sqrt{T})$ regret bound where $K$ is the number of bandits and $T$ is the number of played rounds. Furthermore, we conduct extensive experiments to show that MuFasa outperforms strong baselines on real-world data sets.