Neural Exploitation and Exploration of Contextual Bandits
This work addresses the problem of improving decision-making in contextual bandits for applications like recommendation systems, though it is incremental as it builds on existing neural bandit methods.
The paper tackles the exploitation-exploration trade-off in contextual multi-armed bandits by proposing EE-Net, a neural-based strategy that uses separate networks for exploitation and exploration, achieving an instance-based regret bound of O~(√T) and outperforming baselines on real-world datasets.
In this paper, we study utilizing neural networks for the exploitation and exploration of contextual multi-armed bandits. Contextual multi-armed bandits have been studied for decades with various applications. To solve the exploitation-exploration trade-off in bandits, there are three main techniques: epsilon-greedy, Thompson Sampling (TS), and Upper Confidence Bound (UCB). In recent literature, a series of neural bandit algorithms have been proposed to adapt to the non-linear reward function, combined with TS or UCB strategies for exploration. In this paper, instead of calculating a large-deviation based statistical bound for exploration like previous methods, we propose, ``EE-Net,'' a novel neural-based exploitation and exploration strategy. In addition to using a neural network (Exploitation network) to learn the reward function, EE-Net uses another neural network (Exploration network) to adaptively learn the potential gains compared to the currently estimated reward for exploration. We provide an instance-based $\widetilde{\mathcal{O}}(\sqrt{T})$ regret upper bound for EE-Net and show that EE-Net outperforms related linear and neural contextual bandit baselines on real-world datasets.