dr. Çiçek Güven

Bio

I am an assistant professor in  in the Cognitive Science and Artificial Intelligence department at Tilburg University. My research interests in general are network analysis,  and learning on graph structured data.  I have a mathematical background,  my thesis  involved algebraic graph theory and combinatorics.  I also worked in the industry for a period of four years in banking and consultancy environments as a quant and a data analyst, working with financial and trade data mainly.  Data science projects with a social aspect motivates me a lot.  This interest is effectuated in multiple research projects I am involved in utilising AI to create a more inclusive and accessible world. Some examples are  the 'Child Growth Monitor' by Zero hunger lab;  focusing on malnutrition detection in children,  Ilustre, a project  to support the energy transition efforts in the Caribbean, and the Icon project of Zero poverty lab, linking brain networks and poverty.

Expertise

My research interests are graph theoretical approaches to network analysis with a focus on  structure and dynamics of network evolution,  detecting important substructures possibly for community detection, or explainability purposes and making predictions on graph structural data.   Brain networks or electrical grids can be examples of these networks. I am interested in understanding the link between  the phenomena that are captured in edges- relationships, communication, similarity, exchange of information- and their impact on the structure of the network and formation of motifs and their possible impact on learning.    Network analysis and learning on graph structural data are complementary to each other and a research direction that links these together is  explainability of learning,  in answering questions like what structural properties of the graph leads to certain prediction outcomes?

Courses

• Research Project and Thesis CSAI
• Introduction to Machine Learning
• Master thesis/Data Science in Action
• Bachelorthesis CSAI

Recent publications

1. Image-based body shape estimation to detect malnutrition

   Mohammedkhan, H., Güven, Ç., Balvert, M., & Postma, E. (2023). Image-based body shape estimation to detect malnutrition. Paper presented at Intelligent Systems Conference, Amsterdam.

2. Feature Importance for Clustering

   Nápoles, G., Griffioen, N., Khoshrou, S., & Güven, Ç. (Accepted/In press). Feature Importance for Clustering. In Lecture Notes in Computer Science (LNCS) series Springer.

3. The unique coclique extension property for apartments of buildings

   Brouwer, A., Draisma, J., & Güven, Ç. (Accepted/In press). The unique coclique extension property for apartments of buildings. Innovations in Incidence Geometry — Algebraic, Topological and Combinatorial, 1-16.

4. School Dropout Prediction and Feature Importance Exploration in Malaw…

   Çolak, H., Güven, Ç., & Nápoles, G. (2022). School Dropout Prediction and Feature Importance Exploration in Malawi Using Household Panel Data: Machine Learning Approach. Journal of Computational Social Science.

5. Which is the best model for my data?

   Nápoles, G., Grau, I., Güven, Ç., Özdemir, O., & Salgueiro, Y. (2022). Which is the best model for my data? arXiv.