Bayesian methods for NBC defence

This work demonstrates the use of Bayesian techniques on problems connected with nuclear, biological and chemical (NBC) defence. The overall aim is to determine the most likely characteristics and positions of different types of strikes (including multiple strikes) given imperfect or subjective reports from the field. The problem can be subdivided into two main areas: 1. Identification of the strike parameters (type of agent, quantity of agent, delivery method, etc.); this is a classification problem. 2. Identification of the location of a strike (most likely position and time); this is a tracking problem.