Regret Bounds for Stochastic Shortest Path Problems with Linear Function Approximation
This work addresses a fundamental challenge in reinforcement learning for sequential decision-making under uncertainty, providing a novel algorithmic solution with theoretical guarantees.
The authors tackled the problem of developing efficient algorithms for stochastic shortest path problems using linear function approximation, achieving sublinear regret with computationally efficient stationary policies. This represents the first such algorithm in the linear function approximation literature for stochastic shortest path problems.
We propose an algorithm that uses linear function approximation (LFA) for stochastic shortest path (SSP). Under minimal assumptions, it obtains sublinear regret, is computationally efficient, and uses stationary policies. To our knowledge, this is the first such algorithm in the LFA literature (for SSP or other formulations). Our algorithm is a special case of a more general one, which achieves regret square root in the number of episodes given access to a certain computation oracle.