Maxim Fedotov

Ph.D. student, StatProbML group at Universitat Pompeu Fabra

Maxim Fedotov

I am a Ph.D. student in the Statistics, Probability and Machine Learning Research Group at Universitat Pompeu Fabra under the supervision of David Rossell and Gábor Lugosi.

I have broad research interests including high-dimensional inference, approximate inference and computation, interpretable machine learning, probabilistic guarantees, learning theory in general. I find particularly exciting blending statistical theory and optimization in methods with wide prospects for application.

me

Projects

  • Projected Information Criteria for Variable Selection (with Gábor Lugosi and David Rossell; work in progress)

Presentations

Here is a list of talks and poster presentations I gave.

Contributed talks

  • Projected Information Criteria for Variable Selection
    ICSDS 2025. Seville, Spain. December 15 - 18, 2025.
  • Projected Information Criteria for Variable Selection
    1st INRC 2025. Institute of Mathematics of the University of Seville (IMUS). Seville, Spain. December 11 - 12, 2025.

Campus talks

  • Projected L0 Criteria for Variable Selection in Regression Models
    Lomonosov Moscow State University, Faculty of Economics. Summer, 2025.
  • Classical Analysis of Convergence and Complexity of Non-linear Smooth Optimization Algorithms
    Stats Student Seminar. Universitat Pompeu Fabra, Department of Economics and Business. Fall, 2022.
  • Sequential Monte Carlo for Approximate Variable Selection in Generalized Linear Models
    Stats Student Seminar. Universitat Pompeu Fabra, Department of Economics and Business. Spring, 2022.

Contributed posters

  • Projected and Updated L0 Criteria for Variable Selection in High-Dim. & Large-Sample Regression Models
    Bayes Comp 2025. National University of Singapore. Singapore, Singapore. June 16 – 20, 2025.
  • Projected-L0 Decoder for Variable Selection in Linear Regression
    Mathematical Aspects of Learning Theory – 20 years later. Barcelona, Spain. September 9 – 13, 2024.
  • Projected and One-Step L0 Criteria for Variable Selection
    2024 ISBA World Meeting. Ca' Foscari University of Venice. Venice, Italy. July 1 – 7, 2024.
  • Projected and One-Step L0 Criteria for Variable Selection
    BAYSM 2024. Ca' Foscari University of Venice. Venice, Italy. June 29 – 30, 2024.
  • Sequential Monte Carlo for Approximate Variable Selection in Generalized Linear Models
    Bayes Comp 2023. Levi, Finland. March 15 – 17, 2023.