Using adaptive neural networks for situation recognition in high and low-intensity conflict

Summary form only given, as follows. Tactical military situations, such as battlefield formations and terrorist activities, must first be recognized before effective plans of action can be formulated to counteract or avoid them. Battlefield formations are recognized using visual pattern recognition of the arrangement of military units. Terrorist activities are recognized by frequently occurring associations between concepts such as a person or place. These activities are analyzed over a long period of time to find trends or patterns. An unsupervised neural network was implemented with the adaptive resonance theory paradigm to recognize tactical military situations. The adaptive capabilities were used to overcome the problems of unanticipated threats and events. These networks have successfully identified: battlefield formations indicative of significant tactical situations and frequently occurring events found in reports of terrorist activities