Bayesian Optimization for Formulations Involving Complex Constraints with JMP 19 (2026-EU-30MP-2816)

The production of many products involves mixing or blending multiple ingredients. Finding optimal formulations for these products can be challenging, expensive, and time-consuming, especially if the formulation is subject to complex constraints. The ability to innovate new high-quality products quickly is becoming increasingly more important as technologies advance and competition increases.

We demonstrate how to use various tools in JMP, including the new Bayesian Optimization platform, to speed up and improve the design and analysis of formulation studies in the presence of complex constraints to achieve or maintain a competitive edge. Classical, least squares-based approaches have many subtle difficulties that are almost completely sidestepped by using Gaussian process models and Bayesian optimization. We use case studies that involve complicated constraints, such as restrictions on the number of components allowed in submixtures and mixture of mixtures.