analiza aiqbench
Deep Adaptive Bayesian Screening presents a novel algorithmic framework for optimizing experimental design in high-dimensional discrete spaces, leveraging policy networks to amortize computational costs while integrating Gibbs posterior inference at deployment. By effectively handling sparsity and interactions through specific prior structures, this method offers improved accuracy and scalability over classical baselines under constrained budgets, marking a significant advancement in machine learning-driven optimization techniques.
Deep Adaptive Bayesian Screening
arXiv:2607.16927v1 Announce Type: cross Abstract: We introduce Deep Adaptive Bayesian Screening (DABS), a method for performing adaptive factorial screening in high-dimensional discrete design spaces. DABS learns a policy network offline to sequentially select informative experiments, amortizing Bayesian Optimal Experimental Design. It handles binary designs, incorporates sparsity and interactions via a spike-and-slab prior with strong heredity. The model is trained using a contrastive lower bound on information about factor activity with nuisance effect sizes and noise variance analytically integrated out. Unlike prior amortized Bayesian design approaches, DABS also integrates Gibbs posterior inference at deployment, yielding posterior probabilities of factor activity and credible intervals on effect sizes. We demonstrate DABS on screening problems calibrated to real-world benchmarks and show it achieves superior accuracy and scalability over classical and Bayesian baselines under tight experimental budgets.
Abstract:We introduce Deep Adaptive Bayesian Screening (DABS), a method for performing adaptive factorial screening in high-dimensional discrete design spaces. DABS learns a policy network offline to sequentially select informative experiments, amortizing Bayesian Optimal Experimental Design. It handles binary designs, incorporates sparsity and interactions via a spike-and-slab prior with strong heredity. The model is trained using a contrastive lower bound on information about factor activity with nuisance effect sizes and noise variance analytically integrated out. Unlike prior amortized Bayesian design approaches, DABS also integrates Gibbs posterior inference at deployment, yielding posterior probabilities of factor activity and credible intervals on effect sizes. We demonstrate DABS on screening problems calibrated to real-world benchmarks and show it achieves superior accuracy and scalability over classical and Bayesian baselines under tight experimental budgets.
Machine Learning (stat.ML); Machine Learning (cs.LG)
arXiv:2607.16927 [stat.ML]
(or arXiv:2607.16927v1 [stat.ML] for this version)
https://doi.org/10.48550/arXiv.2607.16927 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 224 cuvinte
Subiecte conexe
Online Learning via Learned Latent Bayesian Tracking
arXiv · 28 sept. 2026
BRAIN: Bayesian Reasoning via Active Inference for Agentic and Embodied Intelligence in Mobile Networks
arXiv · 9 iun. 2026
Sparse Bayesian Deep Functional Learning with Structured Region Selection
arXiv · 5 iun. 2026
Presentation: Engineering AI for Creativity and Curiosity on Mobile
InfoQ · 21 iul. 2026