Juan Kuntz
Juan Kuntz

Senior Research Engineer

About Me

I am a Senior Research Engineer at Polygeist interested in all things machine learning, statistics, data science, optimization, and Monte Carlo. In the near past, I developed tools for machine learning, Bayesian inference, and computational statistics by combining concepts and techniques from optimization and Monte Carlo; see here for a recent example. In the more distant past, I worked on Markov processes, numerical methods for the analysis thereof, and applications of control theory to synthetic biology. I have also been writing a book on Markov chains, see here for the latest draft.

Interests
  • Machine Learning
  • Generative Modelling
  • Computational Statistics
  • Optimization
  • Monte Carlo
Education
  • PhD Bioengineering and Mathematics

    Imperial College London

  • MEng Biomedical Engineering

    Imperial College London

Talks
Publications
(2024). Error bounds for particle gradient descent, and extensions of the log-Sobolev and Talagrand inequalities. Under review.
(2024). Momentum Particle Maximum Likelihood. In ICML 2024.
(2024). The divide-and-conquer sequential Monte Carlo algorithm: theoretical properties and limit theorems. The Annals of Applied Probability.
(2022). Product-form estimators: exploiting independence to scale up Monte Carlo. In Statistics and Computing.