Mining Bayesian Networks to Forecast Adverse Outcomes Related to Acute Coronary Syndrome

摘要: One fascinating aspect of tool building for datamining is the application a generalized to specific domain. Often times, this process results in cross disciplinary analysis both technique and domain itself. This cross-disciplinary often leads not only improvements tool, but more importantly, better understanding underlying model experts involved. paper presents applying identifying Bayesian Network represent dataset triage information taken from patients arriving at emergency room with symptoms Acute Coronary Syndrome. Specifically, expert generated mined Network, trained using dataset, are compared their accuracy forecasting 30-day adverse outcomes represented dataset. The comparison, done ROC curves, shows that Networked slightly outperformed network. discussed direction future work based on complexity network versus expert’s presented..