Using a Baysian belief network to ensure data integrity

摘要

A system assess the risks and/or exposures associated with financial transactions using various statistical and probabilistic techniques. Specifically, plausible sources of error in data used as input to financial risk assessment systems are identified using Bayesian belief networks as a normative diagnostic tool to model relationships between and among inputs/outputs of the risk assessment system and other external factors.