Dynamic feature selection (DFS) based Data clustering technique on sensory data streaming in eHealth record system

Stream clustering in healthcare industry can carry significant importance by discovering disease patterns or by providing better clinical supports. Online stream clustering has several applications associated with it like news filtering, ad filtering, and topic detection. However, clustering particularly for health care industry has not come into consideration yet. In addition, existing clustering methods rarely consider the variety of continuous data and may lead to unsatisfactory results. As a result, implementing existing stream clustering for healthcare industry may not be sustainable for the long run. Motivated from the problem, we propose a clustering algorithm for sensory data in healthcare organisation based on dynamic feature selection known as PCEHRClust. Using a qualitative analysis we show that PCEHRClust is a suitable algorithm for health care industry.