Learning First-Order Symbolic Representations for Planning from the Structure of the State Space
Preprint
Abstract
One of the main obstacles for developing flexible AI systems is the split between data-based learners and model-based solvers. Solvers such as classical planners are very flexible and can deal with a variety of problem instances and goals but require first-order symbolic models. Data-based learners, on the other hand, are robust but do not produce such representations. In this work we address this split by showing how the first-order symbolic representations that are used by planners can be learned from non-symbolic inputs that encode the structure of the state space. The representation learning problem is formulated as the problem of inferring planning instances over a common but unknown first-order domain that account for the structure of the observed state space. This means to infer a complete first-order representation (i.e. general action schemas, relational symbols, and objects) that explains the observed state space structures. The inference problem is cast as a two-level combinatorial search where the outer level searches for values of a small set of hyperparameters and the inner level, solved with SAT, searches for a first-order symbolic model. The framework is shown to produce general and correct first-order representations for standard problems like Gripper, Blocksworld, and Hanoi from input graphs that encode the state-space structure of a single instance.Keywords
All Related Versions
This publication has 15 references indexed in Scilit:
- Reconciling deep learning with symbolic artificial intelligence: representing objects and relationsCurrent Opinion in Behavioral Sciences, 2019
- A review of learning planning action modelsThe Knowledge Engineering Review, 2018
- Human-level intelligence or animal-like abilities?Communications of the ACM, 2018
- Action Schema Networks: Generalised Policies With Deep LearningProceedings of the AAAI Conference on Artificial Intelligence, 2018
- Classical Planning in Deep Latent Space: Bridging the Subsymbolic-Symbolic BoundaryProceedings of the AAAI Conference on Artificial Intelligence, 2018
- Theoretical Impediments to Machine Learning With Seven Sparks from the Causal Revolution2018
- Learning action models from plan examples using weighted MAX-SATArtificial Intelligence, 2007
- Learning Generalized Policies from Planning Examples Using Concept LanguagesApplied Intelligence, 2004
- Learning action strategies for planning domainsArtificial Intelligence, 1999
- The symbol grounding problemPhysica D: Nonlinear Phenomena, 1990