I am currently working as a research scientist at Leibniz Institute ATB in Potsdam. I am part of the Data Science in Bioeconomy department, and also the Understandable Machine Intelligence (UMI) Lab., working on explainable AI.

Before joining ATB, I spent 1 year with Fraunhofer, 8 years with Huawei, 2 years with Netaş, and 3 years with ITU MSPR Lab. in reverse chronological order.

I completed my ERCIM ‘Alain Bensoussan’ postdoctoral fellowship at Fraunhofer Institute for Mechatronic Systems Design (IEM), Trusted Machine Intelligence Group. This fellowship program is organized by European Research Consortium for Informatics and Mathematics. I worked on explainable AI with an emphasis on time series models. I was mainly focused on instance-based, post-hoc counterfactual explanations for black-box classification and anomaly detection models.

Previosly, I was an Engineering Manager (in AI Research) & Senior Research Engineer at Huawei Technologies Turkey R&D Center. At Huawei, I worked on research projects on machine learning, computer vision, video processing, information retrieval, and learning-to-rank. I also gained some experience in grant applications, patent applications, project management, technical team management, and people management.

Before Huawei, I was a Signal Processing Software Design Engineer at Netaş Telecommunications in Istanbul where I developed embedded signal processing software for satellite navigation.

I got my PhD degree in Electrical-Electronics Engineering from Istanbul University-Cerrahpasa (which was formerly part of Istanbul University). My main research topics are machine learning, signal processing, probabilistic modeling, and computer vision. One research direction I am interested in is trustworthy machine learning via explainable AI and modeling/quantifying uncertainties. Some other keywords are Bayesian inference, probabilistic graphical models, subspace learning and sparse approximations, and image/video/audio processing. I am also interested in causal inference and Bayesian statistics (especially approximate inference, sampling, and likelihood-free inference) though I have a limited understanding in these.

During my PhD I worked on probabilistic models for spectral-spatial classification of hyperspectral images under the supervision of Prof. Dr. Aydın Akan and Prof. Dr. Koray Kayabol. In this problem, a land cover segmentation map is obtained by pixel-wise classification. I developed several generative and discriminative classification models with non-stationary mixture models, random fields, deep learning, subspace learning, and optimization algorithms for machine learning.

I pursued my MSc degree from Istanbul Technical University in Biomedical Engineering under the supervision of Prof. Dr. Bilge Günsel. My thesis was about tissue density classification in mammographic images using local features. During master’s studies I was a member of the Multimedia Signal Processing and Pattern Recognition Lab. of ITU, and I took part in several projects on image/video modeling, processing, classification, retrieval, fingerprinting, and copy detection.

I got my bachelor’s degree in Electrical-Electronics Engineering with an emphasis on telecommunications and signal processing from Istanbul University. Graduation project was real-time music and audio effects on DSP.

Some links: