Related Concept Videos
Steps in Outbreak Investigation
Prediction Intervals
However, the point estimate is most likely not the exact value of the population parameter, but close to it. After calculating point estimates, we construct interval estimates, called confidence intervals or prediction intervals. This prediction interval comprises a range of values unlike the point estimate and is a better predictor of the observed sample value, y.
Residuals and Least-Squares Property
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
Applications of GIS: Disaster Management and Emergency Response
The Availability Heuristic
Probability Histograms
You might also read
Related Articles
Articles linked to this work by shared authors, journal, and citation graph.
Organic Chemistry as a Catalyst for AI Innovation: Challenges, Methods, and Emerging Paradigms.
Embedding social determinants in mHealth for pediatric oncology: co-designing a patient-centred tool for febrile neutropenia in resource-limited settings.
Related Experiment Video
Updated: Sep 6, 2025

Trajectory Data Analyses for Pedestrian Space-time Activity Study
Published on: February 25, 2013
Predicting terrorist attacks in the United States using localized news data.
Steven J Krieg1, Christian W Smith2, Rusha Chatterjee2
1Lucy Family Institute for Data and Society, University of Notre Dame, Notre Dame, IN, United States of America.
Machine learning models can predict terrorist attacks using localized news data. A Random Forest model achieved significant accuracy, demonstrating the value of treating attacks as independent events for improved counter-terrorism efforts.
Area of Science:
- Computational Social Science
- Artificial Intelligence
- National Security
Background:
- Terrorism poses a significant global threat, causing numerous fatalities and substantial economic damage annually.
- Existing approaches to terrorism prediction often struggle with sparse and dissimilar historical event data.
- The need for effective predictive models to mitigate terrorism risks is paramount.
Purpose of the Study:
- To develop and evaluate a novel feature representation method for predicting terrorist attacks.
- To assess the efficacy of machine learning models trained on localized news data for terrorism prediction.
- To determine if treating terrorism as independent events improves predictive accuracy.
Main Methods:
- Proposed a novel feature representation method for time-series data.
- Evaluated machine learning models, including a Random Forest with a variable-length moving average.
- Trained and tested models on localized news data for specific states in the United States.
- Utilized the area under the receiver operating characteristic (AUROC) curve to measure model performance.
Main Results:
- The best performing model, a Random Forest, achieved an AUROC of ≥ 0.667 (p ≤ .0001) in four of the five most-impacted states.
- The model's success indicates that localized news data contains valuable predictive information.
- Treating terrorism as a set of independent events proved more effective than a continuous process model.
- Prediction success was not correlated with specific attack characteristics like group or weapon type.
Conclusions:
- Localized machine learning models trained on news data can effectively predict terrorist attacks.
- The approach of modeling terrorism as independent events is beneficial, particularly with limited historical data.
- Future counter-terrorism strategies can leverage these findings for enhanced threat assessment and prevention.
- The study provides a foundation for applying machine learning to national security challenges.

