Computational and applied optimization, systems design, artificial intelligence, machine learning, multi-criterion optimization and decision analysis, and data analytics. The day to day work of the research group lives at the COIN Laboratory.
The reference-point based non-dominated sorting algorithms for multi- and many-objective optimization, cited over 74,000 times combined.
A real-parameter crossover operator that became a default building block across the evolutionary computation field.
Automatically extracting human-interpretable design principles from the Pareto-optimal solutions an optimization run produces.
Evolutionary methods for nested, leader-follower optimization problems that arise across engineering and economics.
Innovized Progress operators, surrogate-assisted search, and learning-based decision-making for expensive optimization problems.
Nonlinear decision trees and explainable policies for autonomous systems, developed with evolutionary optimization.
Current projects, group members, software, and datasets from the Computational Optimization and Innovation Laboratory are maintained on the lab's own site.
