Erdogan Taskesen bio photo

Erdogan Taskesen

Data Scientist in Statistical Machine Learning

Email Google Scholar LinkedIn Github Stackoverflow Orcid Medium Gumroad

Erdogan Taskesen

I work at the Ministry of Infrastructure and Water Management of the Netherlands. The work I do is partly on the (statistical) development, and applications with machine learning but also on Business Intelligence. Before, I worked in the cybersecurity sector and as Data Science consultant (BigDataRepublic) in among others the financial (ING), telecom (KPN), flight (KLM).

During my academic years I worked as a post-doctoral researcher at Delft University of Technology, and later at the VU. I worked on the statistical development of methods for the analysis of cancer genomics and the genetics of Neuroscience. My Ph.D. was at the Erasmus University Medical Center, in colloboration with Weill Cornell Medicine where I developed methods to refine patient stratification in Acute Myeloid Leukemias.

My interest are especially in the statistics behind machine learning, and customizing models to let it work in real-world applications. As an example, we use deeplearning techniques for the detection of objects using satellite images or detection of trucks with dangerous goods. Currently, I am working on explainable AI, causality models, and cyber-security.

Also check out my latest paper at arXiv, HNet: Hypergeometric networks, which is about association learning.