Martijn G枚sgens wins Van Zwet Award for outstanding thesis
We are delighted to announce that Martijn G枚sgens has received the Willem R. van Zwet Award. Presented annually by the Netherlands Society for Statistics and Operations Research (VVSOR), this award recognizes an outstanding Ph.D. in statistics or operations research from a Dutch university.
We are delighted to announce that Martijn G枚sgens has received the Willem R. van Zwet Award. Presented annually by the Netherlands Society for Statistics and Operations Research (VVSOR), this award recognizes an outstanding Ph.D. in statistics or operations research from a Dutch university.
Networks, found in areas like social media, finance, and biology, consist of interconnected elements such as people, companies, or genes. These networks often contain communities, or groups that are more tightly connected to each other than to the rest of the network. Identifying these communities is key to understanding systems and making better decisions. However, many existing methods either group things incorrectly or break them apart.
To solve this, PhD researcher Martijn G枚sgens developed a new approach using a mathematical model to improve community detection and ensure accurate identification at the right level of detail. He defended his thesis on Tuesday, June 11, 2024. For his work, Martijn was awarded the prestigious Willem R. van Zwet award on Thursday, March 20.
The challenge of detecting communities in networks
G枚sgens鈥� research tackled a key challenge in community detection: how to accurately identify groups within a network without over-simplifying or breaking them down into too many smaller parts. Traditional methods often didn鈥檛 handle the complexities of different types of networks well, leading to inconsistent results. Martijn鈥檚 solution was to think of community detection in a new way: using a geometric model that treated different community structures as points in a high-dimensional space. This allowed for a clearer comparison of methods and provided the flexibility to adjust them to find the right balance between large and small communities.
Revealing hidden biases in common algorithms
In addition to improving existing methods, Martijn also pointed out that many common algorithms had hidden biases that made certain types of community structures more likely to be detected. These biases could lead to inaccurate results if not properly addressed. By making these biases clear, Martijn offered a way to refine the algorithms, ensuring they worked better for different kinds of networks and applications.
Real-world impact across multiple fields
G枚sgens鈥� research has wide-ranging potential, improving how networks were analyzed in areas like social media, healthcare, and finance. By making community detection more accurate, G枚sgens鈥� work helped uncover important patterns in complex systems, leading to better decision-making and more reliable results in various fields.
Congratulations to Martijn and the SPOR cluster
What an achievement for Martijn and his supervisors Remco van der Hofstad and Nelly Litvak! This is not the first time a PhD researcher from the SPOR cluster wins the Willem R. van Zwet award. Colin Drent and have been the winners of 2023 and 2024 respectivelly.
Title of PhD thesis:
Supervisors: prof.dr. R.W. van der Hofstad, prof.dr. N.V. Litvak