Detecting and Describing Disease Outbreaks Using Bayesian Networks

Abstract

Disease outbreak detection systems for early detection monitor emergency data for irregularities by comparing the distribution of recent data against baseline distri- bution. In this work, a Bayesian model based on disease outbreak system is used to detect the epidemic. The detection of anthrax epidemic refers to detecting pos- sibility of anthrax attack in minimum number of days. Features related to anthrax disease are taken to develop a probabilistic model using bayesian network. Then by exploring bayesian network we became able to extract the features or patterns that were di erent to common knowledge captured by bayesian model. We imple- mented a rule-based technique that compares recent health-care data against data from a baseline data and nds subgroups of the recent data which shows trend. We experimentally proved that this approach give the detection time of less than 72 hours.