Targeting Insurgent Networks
Clustering attack patterns to pinpoint IED origins in minutes.
Comprehensive Feature Analysis
Incorporated over 300 technical and tactical features for detailed attack series identification.
The Opportunity
Improvised Explosive Device (IED) attacks in Afghanistan posed a significant threat to military operations, requiring timely identification of attack series to locate insurgent networks. Traditional methods of pre-analysis for geoprofiling were manual, time-consuming, and inefficient.
The Approach
Using a nearest neighbor hierarchical clustering algorithm, I analyzed technical and tactical data from attacks to identify connected series of IED incidents. This innovative method reduced analysis time from days to minutes, streamlining the pre-analysis process for geoprofiling.
The Impact
The approach significantly improved counter-IED efforts by enabling precise, efficient identification of attack series. It became the preferred pre-analysis technique for geoprofiling, enhancing operational readiness and contributing to safer deployments.
Features
Time and Spatial Awareness
Factored in geography and temporal patterns to enhance clustering precision.
TF-IDF Style Weighting
Applied intelligent feature weighting to prioritize key attack characteristics effectively.
Rapid, Accurate Results
Achieved near-identical accuracy to three-day analysis in just minutes.
Technology
- Coded in Python (Orange DM)
- MS SQL database
- Lots of coffee – available at the Tim Horton’s in Kandahar 🙂
Disclaimer
The demo images presented here are for illustrative purposes only and use fictitious incidents. No real incidents or sensitive data are shown.
