Learning and Policy Search in Stochastic Dynamical Systems with Bayesian Neural Networks
This work addresses the challenge of capturing complex statistical patterns like multi-modality and heteroskedasticity in transition dynamics for reinforcement learning, which is incremental as it builds on existing Bayesian neural network methods.
The authors tackled the problem of model-based reinforcement learning in stochastic dynamical systems by developing an algorithm that combines Bayesian neural networks with random roll-outs and stochastic optimization, resulting in improved performance on a challenging benchmark and promising results in controlling a gas turbine.
We present an algorithm for model-based reinforcement learning that combines Bayesian neural networks (BNNs) with random roll-outs and stochastic optimization for policy learning. The BNNs are trained by minimizing $α$-divergences, allowing us to capture complicated statistical patterns in the transition dynamics, e.g. multi-modality and heteroskedasticity, which are usually missed by other common modeling approaches. We illustrate the performance of our method by solving a challenging benchmark where model-based approaches usually fail and by obtaining promising results in a real-world scenario for controlling a gas turbine.