Dimos Tsouros, PhD
I am a researcher in AI, with my research interests including, but not being limited to, the following: Constraint Programming, Machine Learning, Artificial Intelligence, Combinatorial Optimization.
I currently am a Post-Doc researcher at KU Leuven, in the Declarative Languages and Artificial Intelligence (DTAI) section of Computer Science, collaborating with Prof. Tias Guns.
I studied Informatics and Telecommunications Engineering at the University of Western Macedonia, Kozani, Greece. After receiving my diploma, I continued doing my Ph.D. on the integration of Constraint Programming and Machine Learning for assisting in the modeling process using learning techniques, with my supervisor being Prof. Kostas Stergiou. I received my Ph.D. from the Department of Electrical and Computer Engineering of the University of Western Macedonia, in 2021. During my Ph.D., I worked as a researcher on several research projects.
Research Interests and Expertise
My research interests include diverse fields of research in Artificial Intelligence, being mainly focused in Constraint Programming and Machine Learning. I am highly interested in constraint-based reasoning techniques for solving hard combinatorial problems in AI, while also being intrigued by the capabilities of learning in AI. Importantly, I am interested in the integration of the reasoning and learning paradigms, having a strong interest on modern constraint solving approaches that learn from the user and the enviroment and are able to explain the decisions made. I have a strong interest on rich interactive techniques between users and solvers, focusing on a more human-aware approach to the solving process.
My main expertise is on interactive constraint acquisition, which involves developing systems that can learn constraints through interaction with (human) users, aiming to widen the use of constraint programming in real-world applications by alleviating the need of expertise in mathematical modeling of the problem constraints. My work in constriant acquisition during my PhD earned an honorable mention in the Doctoral Dissertation Award of ACP in 2022
I also work on explainable constraint solving, where the aim is to explain various aspects of constraint (optimization) problems, regarding why some decisions where made, why there is no solution in some problem or what needs to be changed in order to obtain a better solution. The highlight of my work in explainable constraint solving was the Tutorial presented in CP2023 conference which I co-authored.
Furthermore, I am contributing in the CPMpy modeling library for python, which aligns with my research interests. That is because CPMpy enables easy integration of machine learning and constraint programming, additionally including many state-of-the-art explanation tools for constraint solving.
Besides the above, I have also worked in research projects about real-world applications of machine learning methods. During my PhD, I worked in a project for smart farming and precision agriculture, using machine learning for predictions in UAV imagery. Additionally, for my master thesis, I developed new ML algorithms that yielded variations of random forests and classification via clustering, which were then successfully applied to medical diagnostics