JADAI: Jointly Amortizing Adaptive Design and Bayesian Inference

Bracher, Niels, Kühmichel, Lars, Ivanova, Desi R., Intes, Xavier, Bürkner, Paul-Christian, Radev, Stefan T.

arXiv.org Machine Learning 

We consider problems of parameter estimation where design variables can be actively optimized to maximize information gain. To this end, we introduce JADAI, a framework that jointly amortizes Bayesian adaptive design and inference by training a policy, a history network, and an inference network end-to-end. The networks minimize a generic loss that aggregates incremental reductions in posterior error along experimental sequences. Inference networks are instantiated with diffusion-based posterior estimators that can approximate high-dimensional and multimodal posteriors at every experimental step. Across standard adaptive design benchmarks, JADAI achieves superior or competitive performance.

Duplicate Docs Excel Report

Title
None found

Similar Docs  Excel Report  more

TitleSimilaritySource
None found