Dynamic Bayesian Network Approach for Modeling Trihalomethanes from Ontario Water Supply Systems

Abstract
A dynamic Bayesian network (DBN) approach is used to quantify relational knowledge for modeling relations or dependencies between variables from the dynamic system of disinfection byproduct (DBP) formation which changes over time. The DBN framework is used to assess causality between constituent parameters of water supply quality, based on data from communities in Ontario which rely on groundwater as their source of supply. The DBN models are used to assess probabilistic dimensions and to assist decision-making by identifying control options to decrease DBP formation.
This paper is only available in PDF Format: View full text PDF