We will design “BlenDS” a framework for an intuitive specification of the design space for blends of components. The framework will be based on the following concepts: there are two entities, the “Component” and the “Blend”; they can be recursively combined in a tree-like structure to create the final tree branching from the root blend. The root blend will have all the constraints needed to generate trials of a design space. The scope of this project is to develop an API, use Large Language Models to traslate human written specifications into a root blend, and finally generate valid samples from the blend space that can be then used as a designed experiment. We will compare this easy-to-use approach (whose generation of samples resembles a model-agnostic space filling design), with the classical Mixture Design DOE.

Check out our social media post on LinkedIn!

References:

• NIST, Engineering Statistics Handbook: 5.5.4 What is a Mixture Design?