Military Applications of Data Analytics

Edited by Kevin Huggins

Bringing together a collection of shorter works from experts in the field, this edited book uses a series of case-studies to introduce the concepts and potential benefits of applying data analytic techniques to real-world military problems. Edited by a retired military officer turned university professor of computer science and analytics, though academic in nature, this book demystifies several advanced topics using highly relatable and digestible scenarios.

From the outset, a chapter entitled ‘Bayesian Networks for Descriptive Analytics in Military Equipment Applications’, the subject matter can appear a little daunting. But, also from the outset, the articles are written in a way that develops understanding of the analytic techniques and procedures with little to no discussion of the underlying mathematics. This is proven by the opening chapter, which explains how to create an automated tool for diagnosing generator faults using a software-based decision tree. It outlines how to tackle such a project in easy-to-understand steps, from framing the problem (generator fault diagnosis) to the ways in which data analytics experts work with subject matter experts (generator users and maintainers in this example) to develop the best possible product. As a methodology, this is applicable to any problem and sets the reader up for both the case-studies which follow and initiating a data analytics project of their own.

Another excellent chapter tackles the issue of analysing weather for military applications (vitally important for military aviation). Beginning with a discussion on the collection and collation of data from a wide variety of sources, from bespoke organic sensors to publicly available datasets, the issues of data relevance and overload are introduced. This leads to discussion of the algorithms and methods required to make sense of large datasets and the importance of considering effective information dissemination by design (fit for purpose communication systems!). Again, the methodology and underlying tools discussed are applicable to a wide variety of problems and use-cases and are explained in a way to increase both the readers knowledge and, more importantly, understanding.

Other chapters tackle a range of issues from the use of data analytics for evaluating training effectiveness to considerations for using single board computers in military analytics applications. A wide variety of technical concepts are introduced, including predictive and probabilistic modelling, machine learning, natural language processing, and MapReduce for the processing of extremely large data sets. In turn, each concept is explained in an easy to digest manner and related to a real-world example that offers food for thought in future applications.

Those new to data analytics will find this book an excellent introduction built around a world they know and problems to which they can relate. Those with a surface understanding will benefit from some of the demystification of more complex concepts. Not overly simple (some terms may have the reader reaching for a search engine – no bad thing) but described in a manner that the underlying concepts are easily grasped. Armed with the knowledge this book contains, the reader is sufficiently prepared to identify a problem that data analytics can help solve, to articulate the problem in the appropriate language, and to successfully scope and initiate a data analytics project – essentially everything required to exploit the benefits of data analytics without necessarily knowing how to do it. Exactly where Royal Signals personnel should be thinking. Highly recommended.

Capt Richie Forde

ISBN 978-1-03-247601-8, 194 pages. Available from CRC Press.