Voronoi Tessellations and Machine Learning in Phase Space

by Dr Jamie Gainer (University of Hawaii)

Tuesday Oct 24, 2017, 3:00 PM → 4:00 PM Pacific/Honolulu

112 (WAT)

Cascade decays are prevalent in high energy physics, and are, in particular, a hallmark of many new physics models. The masses of particles in the decay chain can be measured using kinematic endpoints, but this approach has limitations as not all endpoint measurements will be independent, and the measurements may be adversely affected by the experimental resolution. A better approach is to detect the entire multi-dimensional boundary of the region of phase space populated by the cascade decay events. I show how the geometric properties of cells in Voronoi tessellations of event data in phase space, together with machine learning techniques, can be used to detect this boundary and measure the masses of the particles in the decay chain.