US 12,314,870 B2
Dynamically determining a region of effectiveness of athletes during game play based on predictive factors
Aaron K. Baughman, Silver Spring, MD (US); Stefan Van Der Stockt, Austin, TX (US); Craig M. Trim, Ventura, CA (US); John C. Newell, Austin, TX (US); and Stephen C. Hammer, Marietta, GA (US)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by INTERNATIONAL BUSINESS MACHINES CORPORATION, Armonk, NY (US)
Filed on Jun. 28, 2018, as Appl. No. 16/021,799.
Prior Publication US 2020/0005165 A1, Jan. 2, 2020
Int. Cl. G06N 5/04 (2023.01); G06N 20/00 (2019.01)
CPC G06N 5/04 (2013.01) [G06N 20/00 (2019.01)] 24 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
monitoring, by a computing device, real-time sensor data providing information regarding a participant in a sporting event during gameplay of the sporting event;
determining, by the computing device, real-time predictive factors associated with the participant based on the monitoring the real-time sensor data during the gameplay of the sporting event, the real-time predictive factors comprise a trajectory of a ball;
training a machine learning system using machine learning on data which identifies historical effectiveness of play of the participant;
determining weighting factors based on the training data which identifies the historical effectiveness of the play of the participant, wherein the weighting factors comprise a first factor for adjusting a radius for an athlete volume of effectiveness within the training data and a second factor for adjusting a height of the athlete volume of effectiveness within the training data;
determining, by the computing device, a size of a real-time region of effectiveness surrounding the participant during the gameplay of the sporting event based on the predictive factors, the training data identifying historical effectiveness of the play of the participant, and the weighting factors, wherein the real-time region of effectiveness represents a region in which the participant is considered to be effective;
the real-time region of effectiveness surrounding the participant changes based on changes to the real-time predictive factors during the gameplay of the sporting event;
identifying, by the computing device, a set of suggested actions for the participant prior to a play being made by the participant in the sporting event, wherein the set of suggested actions are predicted to minimize a real-time region of effectiveness of a second participant prior to a play being made by the second participant in the sporting event, wherein the second participant is an opponent of the participant;
forecasting, by the computing device, a rate of change of the real-time predictive factors;
determining, by the computing device, a volume of the real-time region of effectiveness for the participant based on the forecasted rate of change of the real-time predictive factors; and
selecting, by the computing device, a shape for the determined volume of the real-time region of effectiveness by utilizing a support vector machine which receives the forecasted rate of change of the real-time predictive factors.