Graphical Models of Intelligent Cause

Graphical models are now widely used to express the underlying mechanisms which drive and explain the processes that actually happen in a given domain. In particular Bayesian Networks and more recently Chain Event Graphs have been used to produce probabilistic predictive models of processes.  Causal algebras are then specified which use this elicited information to determine predictions of what might happen were the system be subjected to various controls. 

But how could we extend this work so that it might apply to produce predictive models of what might happen when the decision maker believes that his controls might be resisted? In this talk I will argue that standard causal models then need to be generalised to embed a decision maker's beliefs of the intent, capability and the information a resistant adversary might have after an intervention. After reviewing recent advances in general forms of Bayesian dynamic causal models I will describe how - using a special form of Adversarial Risk Analysis - we are developing new intelligent algorithms to produce such predictions. The talk will be illustrated throughout by examples of various adversarial threats currently being analysed within the UK. 

This work is part of a wider study by a team based at the Alan Turing Institute and funded by Dstl.