Algorithms

Developing algorithms

Calculations that play a role in decision making often involve optimizing processes. While there are now many algorithms for 'classical' optimization problems, there are great challenges in developing algorithms for problems in a new area. For example, how can computations take into account issues such as fairness or privacy?

Other big questions around algorithms are related to uncertainties and to use of data. How do you include uncertainties in your calculations and models, for example for a complex system like the weather? Another challenge is how to put data into computational models easily and efficiently. Ideally, you would like to make the extraction of information from that data and its use in your calculations, seamless.

These challenges are the focus of algorithm research at CWI. Our goals:

• Design improved algorithms for fundamental optimization problems.
• Discover mathematical techniques that allow you to thoroughly study the quality and behavior of algorithms - so that they can be improved.
• Investigate how uncertainty affects the design and performance of algorithms. We do this, for example, by designing efficient algorithms that evaluate rare high-impact events such as floods or blackouts.
• Investigate how to design and analyze data-informed algorithms, as well as how to implement and use them efficiently. Such algorithms could be used for e.g. climate science, computational fluid dynamics and plasma physics, as well as imaging for healthcare.

You can find more about our research focus areas in the Strategic Plan 2022-2027.