A Machine Learning-Based Approach to Counterinsurgency Strategy with a Novel Game-Theoretic Framework for Performance Evaluation
Defence Science Journal · 27.10.2026 · 76 (6) · 855-869
DOI: 10.14429/dsj.22456 ↗Kurz gefasst
Fachartikel in der Zeitschrift Defence Science Journal, online erschienen am 27.10.2026. Verfasst von Debtoru Chatterjee und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 2.5).
Abstract (Originalsprache)
The RAND Corporation analysed 204 peace-keeping/ counterinsurgency (COIN) operations using 268 COIN attributes. Despite labelling of each counterinsurgency operation as successful (1) or unsuccessful (0), this raw database was not prepared for any machine learning application to predict the probability of success of a particular counterinsurgency mission. The present study converts RAND’s raw data into a CSV file, splits it into training and testing subsets, and applies all the machine learning algorithms on the test set after training. High prediction accuracies exceeding 92 % against the test set lead to the preparation of a front-end user application for predicting the success of any peace-keeping/ counterinsurgency campaign. This application accepts a user input of ‘present’ (1) or ‘absent’ (0) against each specific COIN attribute to rate the probability of success of a peace-keeping operation, using Logistic Regression. The study also proposes a unique, performance metric in place of multiple ones (Accuracy, F-beta score, AUC-ROC curve) as different ML models may outperform each other on differing parameters.
Zitieren
APA
Chatterjee, D., & Kumar, V. (2026). A Machine Learning-Based Approach to Counterinsurgency Strategy with a Novel Game-Theoretic Framework for Performance Evaluation. Defence Science Journal, 76(6), 855-869. https://doi.org/10.14429/dsj.22456
Vancouver
Chatterjee D, Kumar V. A Machine Learning-Based Approach to Counterinsurgency Strategy with a Novel Game-Theoretic Framework for Performance Evaluation. Defence Science Journal. 2026;76(6):855-869. doi:10.14429/dsj.22456
BibTeX
@article{Chatterjee2026,
author = {Chatterjee, Debtoru and Kumar, Vireshwar},
title = {A Machine Learning-Based Approach to Counterinsurgency Strategy with a Novel Game-Theoretic Framework for Performance Evaluation},
journal = {Defence Science Journal},
year = {2026},
volume = {76},
number = {6},
pages = {855--869},
doi = {10.14429/dsj.22456}
}
EUVIMED