Can Quantum Computing Break the “Bellman Curse?”
Science / Computer Science
Stochastic dynamic optimization, a powerful tool for decision-making under uncertainty, faces a formidable challenge when dealing with many variables: the “Bellman curse.” This computational barrier, characterized by exponential growth in complexity, renders traditional approaches impractical for many real-world problems. However, the advent of approximation methods has opened avenues for finding high-quality solutions to specific problems, reigniting interest in overcoming the Bellman curse. The objective of this project is to delve into these methods. We aim to develop new algorithms rooted in reinforcement learning and multiresolution methods while exploring alternative computational paradigms, including concurrent computing and quantum computing.
