U-M Knowledge-Guided Machine Learning (KGML) Workshop

August 4 – August 8, 2025 U-M Knowledge-Guided Machine Learning (KGML) Workshop: Leading the New Paradigm of AI for Science

Location: Michigan League, 3rd Floor, Koessler Room 911 N University Ave Ann Arbor, MI 48109

What is KGML?

Knowledge-Guided Machine Learning (KGML) is a rapidly emerging field that integrates physics-based scientific principles with machine learning algorithms. Unlike traditional machine learning, which relies solely on data, KGML embeds physical laws—such as partial differential equations (PDEs), conservation laws, and symmetries—directly into model architectures, training procedures, or constraints. As the scientific community increasingly recognizes the limitations of purely data-driven ML (uninterpretability, lack of extrapolation, inconsistency), KGML has become a powerful alternative. By weaving domain knowledge into learning workflows, KGML creates more reliable, interpretable, and scientifically grounded AI systems, positioning itself at the forefront of AI for Science (AI4S).

Workshop Overview

This workshop trains participants in the principles and practice of Knowledge-Guided Machine Learning (KGML). This transformative approach blends scientific knowledge—such as physical laws and conservation laws—with machine learning, improving accuracy, interpretability, and trustworthiness, especially for high-stakes scientific applications. The workshop aims to empower fellows to drive the next generation of scientific breakthroughs by integrating domain expertise with machine learning innovation.

Learn more: KGML Workshop Details