Scaling up ML-based Black-box Planning with Partial STRIPS Models
This addresses the challenge for practitioners in sequential decision-making who lack complete models, offering a method to enhance planning without needing more data or architectural changes, though it is incremental.
The paper tackles the problem of improving ML-based black-box planning when a full symbolic model is unavailable, by showing that specifying an incomplete STRIPS model enables the use of relaxation heuristics, leading to effective improvements in planning performance.
A popular approach for sequential decision-making is to perform simulator-based search guided with Machine Learning (ML) methods like policy learning. On the other hand, model-relaxation heuristics can guide the search effectively if a full declarative model is available. In this work, we consider how a practitioner can improve ML-based black-box planning on settings where a complete symbolic model is not available. We show that specifying an incomplete STRIPS model that describes only part of the problem enables the use of relaxation heuristics. Our findings on several planning domains suggest that this is an effective way to improve ML-based black-box planning beyond collecting more data or tuning ML architectures.