two gamma event

High Energy


Joint Theory Seminar: Anders Andreassen


Joint Theory Seminar
Speaker: Anders Andreassen

Title: JUNIPR: a Framework for Unsupervised Machine Learning in Particle Physics

Host:  Verhaaren

Room: 432

Abstract: In applications of machine learning to particle physics, a persistent challenge is how to go beyond discrimination to learn about the underlying physics. In this talk, we will present a new framework: JUNIPR, Jets from UNsupervised Interpretable PRobabilistic models, which uses unsupervised learning to learn the intricate high-dimensional contours of the data upon which it is trained, without reference to pre-established labels. In order to approach such a complex task, JUNIPR is structured intelligently around a leading-order model of the physics underlying the data. In addition to making unsupervised learning tractable, this design actually alleviates existing tensions between performance and interpretability. Applications to discrimination, data-driven Monte Carlo generation and reweighting of events will be discussed. 

Date Mon, January 28, 2019
Time 1:30pm-2:30pm PST
Duration 1 hour
Access Public
Created by High-Energy Seminars
Updated Tue, January 22, 2019 11:37am PST
Send Reminder Yes  -  0 days 4 hour 0 minutes before start

Recent changes RSS feed Creative Commons License Powered by PHP