An AI for Science lab in UConn’s School of Computing, working across materials, chemistry, physics, manufacturing and biomedical engineering
Learning more from less
Scientific data is scarce and expensive. We find the interpretable building blocks of physical systems so that models can learn from small and noisy data, extrapolate beyond it, and tell us in advance when they can’t.
Three ideas run through our work
Decomposing complex systems
We break a system into its components, such as species, reactions, or motifs, and learn how they interact. Recombining components lets our models make predictions on systems they were never trained on.
Decomposing complex predictions
We break a hard prediction into simpler subproblems. Each subproblem is easier to learn, and their solutions can be recombined to answer more complex questions with very little data.
Knowing when it works
We test how well our models extrapolate and quantify their uncertainty. Our goal is to measure whether the training data covers the building blocks a new case needs, so we can evaluate in advance whether recombination will succeed.
We develop new methods in model reduction, symbolic regression, graph neural networks, computer vision, and active learning, driven both by specific scientific applications and by our experience in challenges shared across scientific domains. Our data comes from both simulations and experiments.
About the lab
SciMaLL, led by Qian Yang, is part of the School of Computing at the University of Connecticut. With more than a decade of experience in AI for materials and chemistry, the group brings together researchers from computer science, computational mathematics, and the physical sciences. We work closely with experimental and computational collaborators to build solutions in AI for science that translate to real, measurable impact in the lab.
News
- Aug 2026Qian Yang was promoted to Associate Professor with tenure.
- May 2026Qian Yang received an NSF CAREER Award for Data-Driven Learning of Interpretable and Extrapolative Models of Complex Systems with Sparse Data.
- Mar 2026PhD student Giovanni Romano won first place in the School of Computing at the 12th Annual UConn College of Engineering Graduate Poster Competition.
- Mar 2026Siamese graph neural networks for eutectic melting point prediction was accepted in Digital Discovery.
- Jan 2026DeltaSegment was accepted at IEEE ISBI 2026.
- Dec 2025Congratulations to Dr. Graham Roberts and Dr. Yushuo Niu on completing their PhDs.
- Nov 2025MultiTaskDeltaNet was accepted in Digital Discovery.