A consortium of researchers dedicated to improving the understanding of the human causes and consequences of terrorism

Development and Optimization of Machine Learning Algorithms and Models of Relevance to START Databases

Abstract:

The goal of this project has been to develop an algorithm for finding patterns in terrorism-related databases developed by the National Consortium for the Study of Terrorism and Responses to Terrorism (START), the DHS Center of Excellence at the University of Maryland. The present work expands on the research carried out during the Summer of 2014, with the goal of further developing means to address the missing data problems in the START databases using pattern recognition. In addition, we are working on determining ways to significantly speed up the neural network training process, which can take a fairly long time under our original implementation. While we focused our efforts on specific datasets, this time around we generalized our code so that it could be implemented using any START database.

Publication Information

Full Citation:

Misra, Prabhakar, Raul Garcia-Sanchez, and Daniel Casimir. 2016. "Development and Optimization of Machine Learning Algorithms and Models of Relevance to START Databases." Report to the Office of University Programs, Science & Technology Directorate, U.S. Department of Homeland Security, National Consortium for the Study of Terrorism and Responses to Terrorism (START) Report, University of Maryland, College Park, MD (April). https://www.start.umd.edu/sites/default/files/publications/local_attachments/Prabhakar%20Project%20%232.pdf

START Author(s):