Optimizing Sequential Experimental Design with Deep Reinforcement Learning. (arXiv:2202.00821v3 [cs.LG] UPDATED)

Web: http://arxiv.org/abs/2202.00821

Bayesian approaches developed to solve the optimal design of sequential
experiments are mathematically elegant but computationally challenging.
Recently, techniques using amortization have been proposed to make these
Bayesian approaches practical, by training a parameterized policy that proposes
designs efficiently at deployment time. However, these methods may not
sufficiently explore the design space, require access to a differentiable
probabilistic model and can only optimize over continuous design spaces. Here,
we address these limitations by showing that the problem of optimizing policies …

arxiv deep design experimental learning lg reinforcement reinforcement learning