Experimental Design
Experimental design focuses on optimizing the selection and execution of experiments to maximize information gain while minimizing resource consumption. Current research emphasizes Bayesian approaches, often incorporating machine learning models like neural networks (including normalizing flows) and Gaussian processes, and algorithms such as sequential Monte Carlo and Bayesian optimization, to tackle complex, high-dimensional problems across diverse fields. This improved efficiency in experimental design has significant implications for accelerating scientific discovery and optimizing resource allocation in areas ranging from materials science and drug discovery to robotics and quantum computing.
Papers
Discrete fully probabilistic design: towards a control pipeline for the synthesis of policies from examples
Enrico Ferrentino, Pasquale Chiacchio, Giovanni Russo
Reinforcement Learning based Sequential Batch-sampling for Bayesian Optimal Experimental Design
Yonatan Ashenafi, Piyush Pandita, Sayan Ghosh