Online Planner Selection with Graph Neural Networks and Adaptive Scheduling
This addresses the challenge of improving planning efficiency in AI for researchers and practitioners, though it appears incremental as it builds on existing portfolio-based techniques and graph representations.
The paper tackles the problem of selecting the best automated planner for a given task by proposing a graph neural network (GNN) approach for online planner selection and a two-stage adaptive scheduling method for cost-optimal planning, with experimental results showing effectiveness against strong baselines.
Automated planning is one of the foundational areas of AI. Since no single planner can work well for all tasks and domains, portfolio-based techniques have become increasingly popular in recent years. In particular, deep learning emerges as a promising methodology for online planner selection. Owing to the recent development of structural graph representations of planning tasks, we propose a graph neural network (GNN) approach to selecting candidate planners. GNNs are advantageous over a straightforward alternative, the convolutional neural networks, in that they are invariant to node permutations and that they incorporate node labels for better inference. Additionally, for cost-optimal planning, we propose a two-stage adaptive scheduling method to further improve the likelihood that a given task is solved in time. The scheduler may switch at halftime to a different planner, conditioned on the observed performance of the first one. Experimental results validate the effectiveness of the proposed method against strong baselines, both deep learning and non-deep learning based. The code is available at \url{https://github.com/matenure/GNN_planner}.