Problem
Research questions and friction points this paper is trying to address.
Planar graph generation
Goal-directed generation
Hard constraints
Geometric feasibility
Constrained Goal-directed Planar Graph Generation with Grammar-based Reinforcement Learning
🤖 AI Summary
This study addresses the challenge that existing generative models struggle to produce goal-directed floor plans under strict structural and geometric constraints. We propose a dataset-free constrained Markov decision process formulation that integrates parametric graph grammars with safe reinforcement learning. The core innovation lies in designing a state-dependent safe-set action projection mechanism, which satisfies hard construction constraints in real time while optimizing task objectives. Experimental results demonstrate that our method comprehensively outperforms both traditional and deep generative baselines on a newly introduced benchmark, consistently producing high-quality, feasible floor plan embeddings.
📝 Abstract
Planar graphs are central to applications across science and engineering, yet existing generators provide limited support for goal-directed generation under hard structural and geometric feasibility constraints. We propose a dataset-free method for generating planar graph embeddings by combining parametric graph grammars with safe reinforcement learning to optimize generic task-specific objectives while satisfying constraints during construction. We formulate the generation process as a constrained Markov decision process, where the graph grammar defines the state and action spaces. We further introduce an action projection that maps sampled actions toward state-dependent safe sets, improving constraint satisfaction during training. In contrast to classical graph generators and deep generative models, which typically offer limited goal-directed control or rely on weak constraint satisfaction, our method constructs feasible planar graph embeddings directly during generation. We also introduce a benchmark suite for constrained and goal-directed planar graph generation, together with classical and deep generative baselines. Across all benchmark tasks, our method consistently outperforms baselines while satisfying the formulated constraints.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Planar Graph Generation
Grammar-based Reinforcement Learning
Constrained Markov Decision Process
Safe Reinforcement Learning
Action Projection
🔎 Similar Papers
No similar papers found.
