2019/12/09 by Aaron Havens, Havens, Aaron, Yi Ouyang +5 · 1 citation
Computer Science · Mathematics · #Adversarial Robustness in Machine Learning #Algorithm #Artificial Intelligence (cs.AI) #Artificial intelligence #Computer science #Construct (python library) #Data Stream Mining Techniques #FOS: Computer and information sciences #Latent variable #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine learning #Mathematics #Reinforcement Learning in Robotics #Reinforcement learning #Representation (politics) #Space (punctuation) #State space #State-space representation #cs.AI #cs.LG #stat.ML
paper · pdf · doi:10.48550/arxiv.1912.04201
Deep RL Workshop, Neurips 2019, Vancouver
arxiv created 2019/12/09 · openalex publication_date 2019/12/09 · arxiv updated 2019/12/10 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Model-based reinforcement learning methods typically learn models for high-dimensional state spaces by aiming to reconstruct and predict the original observations. However, drawing inspiration from model-free reinforcement learning, we propose learning a latent dynamics model directly from rewards. In this work, we introduce a model-based planning framework which learns a latent reward prediction model and then plans in the latent state-space. The latent representation is learned exclusively from multi-step reward prediction which we show to be the only necessary information for successful planning. With this framework, we are able to benefit from the concise model-free representation, while still enjoying the data-efficiency of model-based algorithms. We demonstrate our framework in multi-pendulum and multi-cheetah environments where several pendulums or cheetahs are shown to the agent but only one of which produces rewards. In these environments, it is important for the agent to construct a concise latent representation to filter out irrelevant observations. We find that our method can successfully learn an accurate latent reward prediction model in the presence of the irrelevant information while existing model-based methods fail. Planning in the learned latent state-space shows strong performance and high sample efficiency over model-free and model-based baselines.