Questions, methods and evidence
Research
We connect scientific knowledge, data, and simulation to understand complex systems, and examine the assumptions, uncertainty, and evidence behind our conclusions.
AI for Science
How can scientific knowledge, observations, and simulation help uncover mechanisms in complex systems?
Related publications
Research agenda
- Inverse problems and system identification
- Simulation-based surrogate models
- Experimental design
Continuous-Time Modeling
How can we learn, forecast, and simulate continuous dynamics from irregular observations?
Related publications
Research agenda
- Neural differential equations
- Dynamic system identification and forecasting
- Stability and uncertainty
Evidence-Centered AI
What comparisons and uncertainty assessments justify an AI claim?
Related publications
Research agenda
- Validity of comparisons and reproducibility
- Uncertainty assessment
- Distribution shifts and failure analysis
Industrial AI & Decision-Making
How can machine learning, optimization, and simulation support operating decisions in industrial systems?
Related publications
Research agenda
- Quality, reliability, and predictive maintenance
- Resource allocation
- Production and operations planning