Historical operational data analysis for defence preparedness planning. (April 2018)

• Record Type: Journal Article
• Authors: Shafi, Kamran; Debie, Essam; Oliver, David
• Other Names: Kott Alexander guest-editor.
• Is Part Of: Journal of defense modeling and simulation. Volume 15:Number 2(2018:Apr.)
• Page Start: 231
• Page End: 244
• Publication Date: 2018-04
• Subjects: Defence preparedness analysis -- time series analysis -- correspondence analysis -- temporal trend analysis -- artificial neural networks; Military art and science -- Computer simulation -- Periodicals; 355.0011305
• DOI: 10.1177/1548512916664803
• Languages: English
• ISSNs: 1548-5129

Abstract

Preparedness is an important function of defence planning that involves developing defence capabilities to deal with emergent situations relating to national defence and security. Preparedness planning relies on a number of inputs, including requirement analysis, to identify critical capability gaps. Modern data analysis can play an important role in identifying such future requirements. To this end, this paper presents an analytical study, consisting of both descriptive as well as predictive analysis, of historical defence operational data. The descriptive analysis component of the methodology focuses on identifying useful features in the collected data for building a predictive model. The predictive analysis investigates existing patterns in the data, including spatial and temporal trends. An artificial neural network based time series forecasting model is developed to predict future operations based on the identified features. The proposed methodology is applied to a defence operational data set, built from a number of unclassified sources relating to the historical operational deployments of the Australian Defence Force between 1885 and 2012. Implications are also discussed.