BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20251104T002026EST-3623XIoTvt@132.216.98.100
DTSTAMP:20251104T052026Z
DESCRIPTION:Title:Possession Sketches: Mapping NBA Strategies\n\n\n	Abstract
 :\n\n\nWe present Possession Sketches\, a new machine learning method for 
 organizing and exploring a database of basketball player-tracks. Our metho
 d organizes basketball possessions by offensive structure. We first develo
 p a model for populating a dictionary of short\, repeated\, and spatially 
 registered actions. Each action corresponds to an interpretable type of pl
 ayer movement. We examine statistical patterns in these actions\, and show
  how they can be used to describe individual player behavior. Leveraging t
 his vocabulary of actions\, we develop a hierarchical model that describes
  interactions between players. Our approach draws on the topic-modeling li
 terature\, extending Latent Dirichlet Allocation (LDA) through a novel rep
 resentation of player movement data which uses techniques common in animat
 ion and video game design. We show that our model is able to group togethe
 r possessions with similar offensive structure\, allowing for efficient se
 arch and exploration of the entire database of player-tracking data. We sh
 ow that our model finds repeated offensive structure in teams (e.g. strate
 gy)\, providing a much more sophisticated\, yet interpretable lens into ba
 sketball player-tracking data. This is joint work with Andrew Miller.\n
DTSTART:20180921T133000Z
DTEND:20180921T141500Z
LOCATION:Room 001\, Bronfman Building\, CA\, QC\, Montreal\, H3A 1G5\, 1001
  rue Sherbrooke Ouest
SUMMARY:Luke Bornn (Simon Fraser University )
URL:/mathstat/channels/event/luke-bornn-simon-fraser-u
 niversity-289857
END:VEVENT
END:VCALENDAR