Ivan Ovinnikov
Ivan Ovinnikov

Researcher

About Me

I am a researcher in reinforcement learning and generative modeling recently graduated with a doctoral degree from ETH Zürich.

Download CV
Interests
  • Artificial Intelligence
  • Reinforcement Learning
  • Generative Modeling
Education
  • PhD Computer Science (ML/AI)

    ETH Zürich Department of Computer Science

  • MSc Electrical Engineering (Intelligent Systems)

    ETH Zürich Department of Electical Engineering

  • BSc Electrical Engineering

    ETH Zürich Department of Electical Engineering

📚 My Research

I’m an independent researcher working in the intersection of reinforcement learning and generative modeling.

I’m broadly interested in mathematically rich problems applied to this domain.

Please reach out to collaborate 😃

Recent Publications
(2024). Fundamentals of Arthroscopic Surgery Training and beyond: a reinforcement learning exploration and benchmark. International Journal of Computer Assisted Radiology and Surgery.
(2023). Regularizing Adversarial Imitation Learning Using Causal Invariance. arXiv preprint arXiv:2308.09189.
(2019). Poincar$backslash$'e wasserstein autoencoder. arXiv preprint arXiv:1901.01427.
Recent & Upcoming Talks
Recent News

PhD examination

Successfully passed my PhD examination

Teaching

I have been part of a number of courses over the years as a teaching assistant:

  • Advanced Machine Learning (ETH Zürich, 2020, 2021, 2023)
  • Statistical Learning Theory (ETH Zürich, 2020, 2021, 2024)
  • Algorithmic Game Theory (ETH Zürich, 2019)
  • Linear Algebra (ETH Zürich, 2017)
  • Informatik I (ETH Zürich, 2017)

I have also supervised a number of student theses:

  • Ege Karaismaloglu (Bachelor Thesis): Abstract Reasoning in Neural Networks
  • Riccardo De Santi (Research in CS Project): Non-Markovian Policies for Unsupervised Reinforcement Learning in Multiple Environments
  • Marcus Vierneisel (Master Thesis): Exploration of Latent Space in Adversarial Imitation Learning
  • Daniel Garellick (Semester Thesis): Reducing Spurious Correlations in Better-than-Demonstrator Imitation Learning
  • Vitaly Dmitriev (MAS): Applications of Regularized Classification Methods to Surgical Skill Assessment
  • Shengdi Chen (Semester Thesis): An Empirical Study of Primal Wasserstein Imitation Learning in Goal-Conditioned RL