Prof. Dr. Semih Cayci

Forschungsinteressen

Mathematische Grundlagen des Maschinellen Lernens, Deep Learning Theorie, Reinforcement Learning, Angewandte Wahrscheinlichkeit, kontinuierliche Optimierung

Lebenslauf

  • Seit 2022: Tenure-Track Juniorprofessor für Mathematik des Maschinellen Lernens, RWTH Aachen University
  • 2020-2022: NSF TRIPODS Postdoctoral Research Fellow, University of Illinois at Urbana-Champaign
  • 2014-2020: Doctor of Philosophy, Department of Electrical and Computer Engineering, the Ohio State University
  • Teaching

  • 11.46100 Stochastic Analysis - Winter 2024/25
  • 11.80020 Mathematical Foundations of Deep Learning - Winter 2023/24
  • 11.50002 Seminar: Mathematical Learning - Summer 2023
  • 11.70010 Mathematical Foundations of Reinforcement Learning - Winter 2022/23
  • News and Announcements

  • 20.06.2024 - There is an available PhD position in theoretical machine learning. The ideal candidate has a strong background in mathematical optimization, probability theory and machine learning. If you are interested, please contact cayci[at]mathc.rwth-aachen.de.
  • 11.04.2024 - Convergence of Entropy-Regularized Natural Policy Gradient with Linear Function Approximation got accepted to SIAM Journal on Optimization.
  • 20.03.2024 - Finite-Time Analysis of Natural Actor-Critic for POMDPs got accepted to SIAM Journal on Mathematics of Data Science.
  • 18.03.2024 - Finite-Time Analysis of Entropy-Regularized Neural Natural Actor-Critic Algorithms got accepted to the Transactions on Machine Learning Research.
  • 19.02.2024 - New preprint Convergence of Gradient Descent for Recurrent Neural Networks: A Nonasymptotic Analysis is now on arXiv.
  • Kontaktmöglichkeiten und Sprechzeiten

    Büro: Raum 202, Pontdriesch 14

    Telefon: +49 241 80-97828

    E-Mail: cayci@mathc.rwth-aachen.de

    Sprechstunden:
    nach Vereinbarung

    RWTH Aachen University

    Lehrstuhl für Mathematik der Informationsverarbeitung

    z.H. Prof. Dr. Semih Cayci

    Pontdriesch 14

    52062 Aachen

    Valid HTML5 · Valid CSS3