Optimization is an act, process, or methodology of making something as fully perfect or effective as possible. Almost everything can be improved, so optimization’s relevance spans to almost every business or process to make it operate more efficiently and effectively.
Optimization employs mathematical models to discover more efficient ways to control and manage systems, ranging from radiation treatments to data centers and power networks. Optimization researchers at WID solve systems-level problems in emerging science and engineering applications by using optimization technologies in an integrated, interdisciplinary, and collaborative fashion. This includes finding solutions to problems that are the most cost-effective or achieve the highest performance under given constraints by maximizing desired elements and minimizing the undesired elements.
Optimization models promise better process planning that can be tied to and offered by social, economic, and financial systems. Certain social and political constraints have caused optimization to go largely unexplored, as have methods for translating plans into policy. We hope to draw on collaborations with communications experts, political scientists, sociologists, economists, behavioral scientists, and business professionals to further leverage optimization’s potential for boosting efficiencies and improving systems that reach into all corners of our lives.
Learn more about Optimization at WID. It is a key component of WID’s Data Science Hub.
By combining information from many farms, predictive models and analytic tools can be developed to help producers and consultants navigate, visualize. and analyze the data they are getting from an increasing number of sources to support better management decisions.
A new approach to climate data analysis hopes to improve regional forecasts.
Investigators from WID are among the recipients of the latest round of UW2020 awards.
Laurent Lessard is working to improve the algorithms and computer software that keep the modern world running smoothly.
Karen Schloss and Laurent Lessard are working on a method for matching colors to people’s expectations to send the right message — starting with the best colors for waste and recycling bins.
Assistant Professor
Polyhedral methods and algorithms for combinatorial problems
The new institute, housed at UW–Madison’s Wisconsin Institute for Discovery (WID), will play a key role in the future of data science, developing fundamental techniques for handling increasingly massive data sets in shorter times.
Assistant Professor
High-dimensional statistics, compressed sensing, nonconvex optimization, robust statistics, and network inference.
Professor and Department Chair
Models and Algorithms for Large-Scale Numerical Optimization
Large scale machine learning, distributed optimization and active learning.
Numerical optimization, especially problems involving real (as opposed to integer or discrete) variables. Includes theory, algorithms, implementation and application.
The NRG focuses on signal processing, machine learning, optimization, and statistics. Areas of focus include sparsity and active learning, learning graphs and networks, and interactive machine learning with humans.
Interests include energy markets, Climate Policy, international trade, technical change and computational economics.
Theoretical and algorithmic aspects of mixed-integer optimization, with a special emphasis in linear and polynomial functions. Other interests include polyhedral combinatorics and combinatorial optimization.
Algorithmic and interface development for large scale problems in mathematical programming, including links to the GAMS and AMPL modeling languages, and general purpose software such as PATH, NLPEC and FATCOP.
Convergence analysis of convex optimization algorithms using techniques from robust control theory
Metric learning and active machine learning, focussed on reducing the sample complexity of learning.
Assistant Scientist
Building optimization models that help untangle issues of environmental sustainability.
Associate Professor
Design of exact and approximate algorithms for mixed-integer optimization problems
