Blended Clustering for Health Data Mining

摘要

Exploratory data analysis using data mining techniques is becoming more popular for investigating subtle relationships in health data, for which direct data collection trials would not be possible. Health data mining involving clustering for large complex data sets in such cases is often limited by insufficient key indicative variables. When a conventional clustering technique is then applied, the results may be too imprecise, or may be inappropriately clustered according to expectations. This paper suggests an approach which can offer greater range of choice for generating potential clusters of interest, from which a better outcome might in turn be obtained by aggregating the results. An example use case based on health services utilization characterization according to socio-demographic background is discussed and the blended clustering approach being taken for it is described.