Roberto Corizzo develops adaptive machine learning methods for evolving data, focusing on continual learning, forecasting, anomaly detection, and class imbalance across healthcare, finance, cybersecurity, and scientific discovery.
 

Johannes Lange develops importance sampling techniques for software that uses neural networks to accelerate statistical inference in astrophysics and cosmology, including Nautilus, a reference open-source python implementation of this technique for Bayesian posterior and evidence estimation.
 

Nathalie Japkowicz designed the Hate Mitigation App for discovering emergent hate speech on social media as part of her Signature Research Initiative interdisciplinary project, Unmasking Coded Hate Speech.

Leah Ding develops trustworthy AI systems for scientific discovery. Her research, funded by the National Science Foundation and NASA, includes a major NSF grant to advance AI and machine learning for wildfire detection and forecasting.
 

Patrick Wu published "Beyond Price: A Technical Quality Framework for AI Antitrust" at the 2026 ACM Conference on Fairness, Accountability, and Transparency, arguing that competition in the AI market should be analyzed through the lens of quality.
 

News & Updates

Michael Robinson received supplemental funding of $50,000 (new total: $110,000) from the Pacific Northwest National Laboratory for “Modeling and Analytic Capabilities for KBase”.

Michael Robinson’s paper, “Token embeddings violate the manifold hypothesis,” which investigates the internal geometry of large language models, was accepted to NeurIPS 2026, one of the world’s top AI conferences. 

Nathalie Japkowicz and ZoisBoukouvalas published Machine Learning Evaluation: Towards Reliable and Responsible AI (Cambridge University Press), which examines ways to evaluate machine learning systems, including fairness and bias.