Data-Driven Discovery of Iterative Algorithms

Iterative algorithms are the basis of scientific computing and are used across chemical engineering to solve process design, simulation, optimization, and control problems. These algorithms are also fundamental in modern machine learning, used to train surrogate models, such as physics-informed neural networks. Despite significant advances, the development of iterative algorithms is a human-intuition-driven and time-consuming task. We develop data-driven algorithms to automate the discovery of iterative algorithms. 

Selected Publications

1. Tongjia Liu and Ilias Mitrai, Symbolic Discovery of Iterative Algorithms: A Continuous Latent Space Bayesian Optimization Framework, Under review, 2026 [arXiv]