I am a PhD candidate at Ecole Polytechnique Fédérale de Lausanne (EPFL) since September 2018 in the department of Computer and Communication Sciences. My primary research goal is to mathematically understand the generalization behavior of some machine learning models including Random Features, Kernel Methods, and Neural Networks. I am very lucky to have my inspiring supervisors Clément Hongler and Wulfram Gerstner.
Before my PhD, I studied Electrical and Electronics Engineering and Mathematics (double-major) at Koç University and I spent half a year at the University of California Los Angeles (UCLA) as an exchange student. I went to Izmir Science Highschool where I competed in Maths Olympiads and earned two International Mathematical Olympiad (IMO) medals. I dance Argentine Tango.
We discuss how the effective ridge reveals the implicit regularization effect of finite sampling in random features. The derivative of the effective ridge tracks the variance of the optimal predictor, yielding an explanation for the variance explosion at the interpolation threshold for arbitrary datasets.