ODSL Forum: Max Lamparth, Gaussian Processes for high precision experiments

Friday 25 Feb 2022, 14:00 → 15:00 Europe/Berlin

Presenter: Max Lamparth, TUM
Title: Gaussian Processes for high precision experiments

Abstract:

Gaussian processes provide a framework to model physical phenomena in a Bayesian way while using little prior assumptions and being robust to overfitting [1]. We use this to model systematic corrections of high-precision experiments to improve our measurement results. 
This talk presents the mathematical background of Gaussian processes with physics applications for regression and Bayesian optimization [2]. We also present advanced Gaussian process models like stochastic variational Gaussian processes with further physics applications [3, 4].

Reference:

[1] C. E. Rasmussen and C. K. I. Williams, Gaussian Processes for Machine Learning, 2nd ed. (The MIT Press,2006).
[2] B. Shahriari, K. Swersky, Z. Wang, R. P. Adams, and N. de Freitas, Taking the human out of the loop: A review of bayesian optimization, Proceedings of the IEEE104,148 (2016).
[3] J. Hensman, N. Fusi, and N. D. Lawrence, Gaussian processes for big data, arXiv preprint arXiv:1309.6835(2013).
[4] D. M. Blei, A. Kucukelbir, and J. D. McAuliffe, Variational inference: A review for statisticians, Journal of the American statistical Association 112, 859 (2017).

As usual, the format will be a short presentation followed by plenty of discussion.

Please let all interested know about the Journal Club (get them to send their Email address).  We are looking forward to lively discussions.