Bayesian optimization with known experimental and design constraints for chemistry applications
Riley J. Hickman, Matteo Aldeghi, Florian Häse, Alán Aspuru‐Guzik
Abstract
A Bayesian optimization algorithm that satisfies known constraints has been developed. The usefulness of considering experimental and design constraints are shown in two simulated chemistry applications.
Topics & Concepts
Computer scienceChemical spaceBayesian optimizationRobustness (evolution)Flexibility (engineering)A priori and a posterioriMathematical optimizationArtificial intelligenceChemistryPhilosophyEpistemologyDrug discoveryBiochemistryStatisticsGeneMathematicsMachine Learning in Materials ScienceProcess Optimization and IntegrationComputational Drug Discovery Methods