Contextual Bandits for Evaluating and Improving Inventory Control Policies
This work addresses inventory management challenges for businesses, offering a practical tool for policy evaluation and incremental improvement.
The paper tackles the problem of evaluating and improving inventory control policies under nonstationary demand, lost sales, and stochastic lead times by introducing an equilibrium policy concept and a contextual bandit algorithm. The result shows that this method achieves favorable theoretical guarantees and empirical performance.
Solutions to address the periodic review inventory control problem with nonstationary random demand, lost sales, and stochastic vendor lead times typically involve making strong assumptions on the dynamics for either approximation or simulation, and applying methods such as optimization, dynamic programming, or reinforcement learning. Therefore, it is important to analyze and evaluate any inventory control policy, in particular to see if there is room for improvement. We introduce the concept of an equilibrium policy, a desirable property of a policy that intuitively means that, in hindsight, changing only a small fraction of actions does not result in materially more reward. We provide a light-weight contextual bandit-based algorithm to evaluate and occasionally tweak policies, and show that this method achieves favorable guarantees, both theoretically and in empirical studies.