Compression of behavioral data using clustering technique

摘要

An efficient method for compression of behavioral data is presented in this paper. The data sets, which occurs most in a wide variety of applications, create some of the most significant challenges in data mining. We propose a method expressing the alikeness of data. Data reduction technique using statistical technique widely known as clustering method is extensively used in management science to reduce the behavioral data. The use of clustering method in data compression is largely unexplored area which has great potential for further research. This study is a humble attempt in this direction. Stating a hypothesis that more number of clusters the less will be the deviation from the original data.