Detecting and Understanding Dysfunctional Anomalies in Queensland Healthcare Databases (2008-2011)
Abstract
Healthcare systems are large complex organisations that are required to function effectively and efficiently. They routinely produce a large amount of data that cannot be comprehensively analysed using manual procedures. This project will develop and apply state-of-the-art data mining and machine learning techniques to support the detection of anomalies in healthcare databases. Improvements in anomaly detection will lead to earlier detection and a better understanding of problems in the system. Techniques will be developed and deployed in case studies on data for hospital bed occupancy, cardiac units and emergency units in Queensland Health.
