Yilmazcan Özyurt

Yilmazcan Özyurt

Yilmazcan Özyurt

Student / Programme Doctorate at D-INFK

ETH Zürich

Professur für Informatik

STF G 222

Stampfenbachstrasse 114

8092 Zürich

Switzerland

Additional information

Research area

Deep Learning, Time Series, Domain Adaptation, Transfer Learning, Generative Models

CV PDF

Additional information

I am a second year Doctoral Student in DS3Lab led by Prof. Ce Zhang. My research focuses on enabling time series models to solve real-world problems in various fields, including healthcare, e-commerce, and social networks. Time series data is generated and processed everywhere, yet we lack the large-scale, global deep learning models in time series (unlike CV or NLP) that could unlock new applications. To understand the challenges, I am interested in modeling the data generating process of time series, which facilitates better generalization over the unseen examples. Further, I have been working on domain adaption to extract the contextual representations of time series, which is transferable across domains (e.g., regions, institutions, people). 

 

Before I started my doctoral degree, I completed my BSc. in Computer Engineering and Industrial Engineering (double major) from Koç University (Turkey). Then I moved to ETH Zürich to complete my MSc. Degree in Computer Science with the focus on machine learning. My MSc. thesis has been published in KDD’21. 

 

Selected papers:

 

Ozyurt, Y., Hatt, T., Zhang, C., & Feuerriegel, S. (2022). "A Deep Markov Model for Clickstream Analytics in Online Shopping". In Proceedings of the Web Conference 2022. (linkcall_made)

 

Ozyurt, Y., Feuerriegel, S., & Zhang, C. (2022). "Contrastive Learning for Unsupervised Domain Adaptation of Time Series". arXivcall_made

 

Ozyurt, Y., Kraus, M., Hatt, T., & Feuerriegel, S. (2021). "AttDMM: An Attentive Deep Markov Model for Risk Scoring in Intensive Care Units." In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining. (linkcall_made)

JavaScript has been disabled in your browser