Applying Data Mining and Machine Learning Techniques to Submarine Intelligence Analysis

Abstract

We describe how specialized database technology and data analysis methods were applied by the Swedish defense to help deal with the violation of Swedish marine territory by foreign submarine intruders during the Eighties and early Nineties. Among several approaches tried some yielded interesting information, although most of the key questions remain unanswered. We conclude with a survey of belief-function- and genetic-algorithm-based methods which were proposed to support interpretation of intelligence reports and prediction of future submarine positions, respectively.