Step-DAD: Semi-Amortized Policy-Based Bayesian Experimental Design
This work addresses experimental design in Bayesian settings, offering a more flexible and robust approach, though it appears incremental as it builds on existing policy-based methods.
The paper tackles the problem of Bayesian experimental design by introducing Step-DAD, a semi-amortized policy-based method that updates the design policy during experiments, resulting in improved decision-making and robustness compared to state-of-the-art methods.
We develop a semi-amortized, policy-based, approach to Bayesian experimental design (BED) called Stepwise Deep Adaptive Design (Step-DAD). Like existing, fully amortized, policy-based BED approaches, Step-DAD trains a design policy upfront before the experiment. However, rather than keeping this policy fixed, Step-DAD periodically updates it as data is gathered, refining it to the particular experimental instance. This test-time adaptation improves both the flexibility and the robustness of the design strategy compared with existing approaches. Empirically, Step-DAD consistently demonstrates superior decision-making and robustness compared with current state-of-the-art BED methods.