LGAIMLJul 11, 2020

Control as Hybrid Inference

arXiv:2007.05838v110 citations
Originality Incremental advance
AI Analysis

This provides a principled framework for combining sample efficiency and asymptotic performance in reinforcement learning, though it appears incremental in unifying existing approaches.

The paper tackles the split between model-based and model-free reinforcement learning by unifying them into a 'control as hybrid inference' (CHI) framework, demonstrating that it outperforms strong baselines on a continuous control benchmark.

The field of reinforcement learning can be split into model-based and model-free methods. Here, we unify these approaches by casting model-free policy optimisation as amortised variational inference, and model-based planning as iterative variational inference, within a `control as hybrid inference' (CHI) framework. We present an implementation of CHI which naturally mediates the balance between iterative and amortised inference. Using a didactic experiment, we demonstrate that the proposed algorithm operates in a model-based manner at the onset of learning, before converging to a model-free algorithm once sufficient data have been collected. We verify the scalability of our algorithm on a continuous control benchmark, demonstrating that it outperforms strong model-free and model-based baselines. CHI thus provides a principled framework for harnessing the sample efficiency of model-based planning while retaining the asymptotic performance of model-free policy optimisation.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes