Deep Generative Models for Decision-Making and Control
This work addresses limitations in model-based reinforcement learning methods, which could improve their practical usefulness for AI and robotics applications, though it appears incremental as it builds on existing approaches.
The thesis tackled the empirical shortcomings of deep model-based reinforcement learning for decision-making and control by studying their causes and proposing solutions, highlighting how generative modeling techniques like beam search and classifier-guided sampling can be reinterpreted as planning strategies.
Deep model-based reinforcement learning methods offer a conceptually simple approach to the decision-making and control problem: use learning for the purpose of estimating an approximate dynamics model, and offload the rest of the work to classical trajectory optimization. However, this combination has a number of empirical shortcomings, limiting the usefulness of model-based methods in practice. The dual purpose of this thesis is to study the reasons for these shortcomings and to propose solutions for the uncovered problems. Along the way, we highlight how inference techniques from the contemporary generative modeling toolbox, including beam search, classifier-guided sampling, and image inpainting, can be reinterpreted as viable planning strategies for reinforcement learning problems.