Landmark-Assisted Monte Carlo Planning
This work addresses the challenge of enhancing anytime algorithms for solving Markov Decision Processes, which is incremental as it extends landmark concepts from classical to stochastic planning.
The paper tackled the problem of using landmarks to improve online probabilistic planning in stochastic domains by formalizing probabilistic landmarks and adapting the UCT algorithm to leverage them as subgoals, resulting in significant performance improvements in benchmark domains, though the optimal balance between greedy landmark achievement and long-term goal achievement varies by problem.
Landmarks$\unicode{x2013}$conditions that must be satisfied at some point in every solution plan$\unicode{x2013}$have contributed to major advancements in classical planning, but they have seldom been used in stochastic domains. We formalize probabilistic landmarks and adapt the UCT algorithm to leverage them as subgoals to decompose MDPs; core to the adaptation is balancing between greedy landmark achievement and final goal achievement. Our results in benchmark domains show that well-chosen landmarks can significantly improve the performance of UCT in online probabilistic planning, while the best balance of greedy versus long-term goal achievement is problem-dependent. The results suggest that landmarks can provide helpful guidance for anytime algorithms solving MDPs.