Applied Visual Analytics for Event Prediction

Back to all technologies
Download as PDF
Researchers at Purdue University have recently developed an innovative software package that analyzes data, visually displays the results, and predicts the location of hot spots or areas where unusually high incidence of events occur. This predictive visual analytics toolkit provides analysts with linked geospatiotemporal and statistical analytic views to facilitate the forecasting of hotspots. The system models spatiotemporal events through the combination of kernel density estimation for event distribution and seasonal trend decomposition by loess smoothing for temporal predictions. Analysts are provided with estimates of error in the temporal modeling along with temporal altering to indicate the occurrence of hotspots. Spatial data is distributed based on a modeling of previous event locations; thereby, maintaining a temporal coherence with past events. The program allows analysts to perform real-time hypothesis testing, plan intervention strategies, and allocate resources to correspond to perceived threats.

-Formulates a prediction about an event
-Presents analysts with a visual representation of data
-Specifically designed for homeland security personnel, first responders, and law enforcement

Potential Applications:
-Computer technology
Nov 28, 2011
United States
Dec 30, 2014

May 28, 2010

May 29, 2009
United States
Purdue Office of Technology Commercialization
1801 Newman Road
West Lafayette, IN 47906

Phone: (765) 588-3475
Fax: (765) 463-3486