Search papers, labs, and topics across Lattice.
This paper introduces a model-driven approach to efficiently develop families of reinforcement learning (RL) environments, addressing the challenges of labor-intensive and error-prone manual processes. By employing a hybrid genetic algorithm that combines global and local search strategies, the authors automate the generation of diverse yet similar training environments through model transformations. The approach is validated in a wildfire mitigation scenario and demonstrates significant improvements in scalability and adaptability for RL training.
Automating the creation of RL environment families could drastically reduce development time and enhance agent training efficiency.
Virtual training environments are software-intensive systems in which reinforcement learning (RL) agents learn, adapt, and demonstrate meaningful behavior. Virtual training environments offer a safe and cost-efficient alternative to training agents in real-world settings. However, to converge, most realistic RL problems require training in multiple, mostly similar but slightly different environments - i.e., families of environment variants. The typical development process of environment families is a labor-intensive and error-prone manual endeavor that does not scale well. To alleviate these issues, in this paper, we propose a model-driven approach for developing families of RL training environments. To obtain the family of environments, we develop an approach and prototype tool. In our approach, a hybrid genetic algorithm - a combination of population-based global search and heuristic local search - generates environment families. Mutations and constraints are expressed as model transformations and are operationalized into a search process by a state-of-the-art model transformation engine. We demonstrate the soundness of our approach in a wildfire mitigation scenario and curriculum learning - a particular learning paradigm that relies on environment families.
