PROBABILISTIC MODEL USING CORRELATES OF SENSITIVE TECHNOLOGIES AND EXPERT JUDGEMENT

Pacific Northwest National Laboratory (PNNL) and Texas A&M University (TAMU) collaborated to develop a Bayesian network model that estimates the likelihood that a state will proliferate. The model is based on statistical investigation performed by academic International Relations community associated with the spatial proximity and geopolitical relationships between states. The model uses compiled data, and ideas developed by Fuhrmann and Tkach (TAMU), and PNNL’s experience in developing Bayesian network models of social behavior. Bayesian network models can be informed by expert judgement as well as historical data, which may be important given that factors associated with whether a state decides to pursue and acquire nuclear weapons capabilities change over time. The primary purpose of this paper is to demonstrate the benefits of performing analysis using a Bayesian network model in assessing state proliferation.