Learning General Policies From Examples
This addresses scalability issues for researchers and practitioners in automated planning who need interpretable and correct policies for large-scale problems.
The paper tackles the scalability limitations of combinatorial methods for learning general planning policies by proposing a new symbolic method based on hitting set algorithms instead of SAT/ASP, enabling it to handle problems with millions of states and hundreds of thousands of features.
Combinatorial methods for learning general policies that solve large collections of planning problems have been recently developed. One of their strengths, in relation to deep learning approaches, is that the resulting policies can be understood and shown to be correct. A weakness is that the methods do not scale up and learn only from small training instances and feature pools that contain a few hundreds of states and features at most. In this work, we propose a new symbolic method for learning policies based on the generalization of sampled plans that ensures structural termination and hence acyclicity. The proposed learning approach is not based on SAT/ASP, as previous symbolic methods, but on a hitting set algorithm that can effectively handle problems with millions of states, and pools with hundreds of thousands of features. The formal properties of the approach are analyzed, and its scalability is tested on a number of benchmarks.