I am a postdoctoral researcher at the KERMIT (knowledge-based systems) group at Ghent University.
Machine intelligence and living systems fascinate me. In my research, I develop intelligent techniques to understand, predict and control biological networks. My main toolbox involves a mix of machine learning, optimization, bioinformatics and graph theory. I use these methods to predict how plants, animals, microorganisms and molecules interact with each other.
Much of my work involves working together with others, translating biological problems as mathematical or computational ones. Every year, I try to engage students students in projects and theses, doing cool things such as making a beer classifier or designing new proteins.
During my years as a teaching assistant, I was involved in various courses on data analytics and computational intelligence, including statistics, probability theory and machine learning. Now, I am the responsible teacher for the course 'Selected Topics in Mathematical Optimization', learning master students of bioinformatics how solve concrete problems.
A general-purpose toolbox for efficient Kronecker-based learning
Pairwise learning is a machine learning paradigm where the goal is to predict properties of pairs of objects. Applications include recommender systems, such as used by Amazon, molecular network inference and ecological interaction prediction. Kronecker-based learning systems provide a simple, yet elegant method to learn from such pairs. Using tricks from linear algebra, these models can be trained, tuned and validated on large datasets. The Julia package Kronecker.jl aggregates these tricks, such that it is easy to build such learning systems.