Many real-world decision-making processes involving finding optimal actions can be formulated mathematically as optimization problems, and solved to provable optimality with existing software (e.g., CPLEX, Gurobi, Google OR-Tools, etc.) or heuristically.
The course provides an introduction to these mathematical models and how to use them in practical decision-making tasks.
The course aims at showing how collected data can be used not only to describe and predict current and future processes, but also to change them so that the usage of resources is minimized and desired goals are
In the course you will go through the following elements:
Overview of available technologies for solving optimization problems: e.g., integer and linear programming, constraint programming and heuristics.
Formulation of decision processes as optimization problems using a formal language, i.e. the model, which can be understood and solved by the solver in the above technologies.
Implementation of models in a modeling language: MiniZinc, CPMpy, or similar.
Participants may (but need not) have an input on which technology will be most focused on.
At the end of the course, you will be able to:
Recognize and formulate real-world decision tasks as optimization problems.
Use data to prescribe better ways to do things.
Know what existing technology can do for solving the real-world optimization problems that arise in your business.
Computer scientists, data scientists, and mathematicians and economists with basic programming skills.
English or Danish.
Associate Professor Marco Chiarandini studies a wide range of issues in mathematical optimization methods for automated timetabling, scheduling and routing with applications in the industry and in the public sector.
His research interests include also the application of statistical learning methods to the analysis and configuration of optimization algorithms.