Dimension and computation reduction approach for K-Means clustering algorithm for Big Data

This paper proposes a method to reduce the computations of the K-Means clustering algorithm for big data. First, with the PCA algorithm, the dimensions of datasets are reduced to one or two dimensions, and then with using the information of distance from one point to its two nearest centers and their changes in the last two iterations lead to an increase of the speed and quality of the K-Means algorithm.Using real samples and experiments, it was ensured that at the best case the speed of the proposed method was improved by ۹۵.۹۱% and the quality of the proposed method was improved by ۹۹.۷۱%. These findings show that the proposed method is very useful for big data.