Our next AI Developer Workshop will take place on April 7th, 2025!
Dr. Johannes Peter Dürholt, Senior Data Scientist at Evonik, will present key aspects of his research in his talk titled “How to plan experiments in the most efficient way using probabilistic machine learning?“.
Efficient experimentation is the cornerstone of success in the chemical industry, driving innovation and reducing time-to-market for new products. In this presentation, Evonik’s journey from classical Design of Experiments (DoE) to Bayesian Optimization (BO) techniques is outlined. We will introduce our Python codebase, BoFire, specifically designed to implement DoE and BO methods tailored for experiments in the chemical domain and beyond. BoFire enables researchers to systematically navigate complex parameter spaces, iteratively refine experimental designs, and make data-driven decisions grounded in probabilistic models both in closed loop and human-in-the-loop optimization scenarios.
We look forward to exciting insights and an engaging discussion.
