HET Brown Bag Seminars | Improving Numerical Integration and Event Generation with Normalizing Flows
Claudius Krause (Fermilab)
With the upcoming HL-LHC, the budget for computing will be insufficient to generate a sufficient amount of Monte-Carlo events for both signal and background predictions. The driving force behind these costs is the inefficiency of the Monte-Carlo phase space generators and the unweighting efficiencies.
After a short review of traditional algorithms, I will introduce a new Machine Learning algorithm that uses Normalizing Flows for efficient numerical integration and random sampling. This approach is especially efficient in high-dimensional integration spaces. I will show some preliminary results obtained with the matrix element generator of Sherpa and discuss different choices of hyperparameters and their influence on the result.
After a short review of traditional algorithms, I will introduce a new Machine Learning algorithm that uses Normalizing Flows for efficient numerical integration and random sampling. This approach is especially efficient in high-dimensional integration spaces. I will show some preliminary results obtained with the matrix element generator of Sherpa and discuss different choices of hyperparameters and their influence on the result.
Building: | Randall Laboratory |
---|---|
Website: | |
Event Type: | Lecture / Discussion |
Tags: | brown bag, Fall 2019, Physics |
Source: | Happening @ Michigan from Leinweber Center for Theoretical Physics, Department of Physics, HET Brown Bag Series, Leinweber Center for Theoretical Physics Seminars, Leinweber Center for Theoretical Physics |