A New Optimal Stepsize For Approximate Dynamic Programming
This work addresses a key bottleneck in ADP algorithms for computationally intensive operations research applications, offering a more robust and efficient solution.
The authors tackled the problem of stepsize selection in approximate dynamic programming, which is crucial for performance but often requires tuning. They derived a new stepsize rule that optimizes prediction error, resulting in faster convergence and reduced sensitivity to tuning in numerical experiments.
Approximate dynamic programming (ADP) has proven itself in a wide range of applications spanning large-scale transportation problems, health care, revenue management, and energy systems. The design of effective ADP algorithms has many dimensions, but one crucial factor is the stepsize rule used to update a value function approximation. Many operations research applications are computationally intensive, and it is important to obtain good results quickly. Furthermore, the most popular stepsize formulas use tunable parameters and can produce very poor results if tuned improperly. We derive a new stepsize rule that optimizes the prediction error in order to improve the short-term performance of an ADP algorithm. With only one, relatively insensitive tunable parameter, the new rule adapts to the level of noise in the problem and produces faster convergence in numerical experiments.