AN INCREASED PERFORMANCE OF CLUSTERING HIGH DIMENSIONAL DATA THROUGH DIMENSIONALITY REDUCTION TECHNIQUE

Abstract

With the incredible growth of high dimensional data such as microarray gene expression data, the researchers are forced to develop some new techniques rather than using existing techniques to meet their requirements. Microarray is a mechanism of measure the expression level of tens of thousands of genes simultaneously as a result, the data generated is very large. There must be an efficient technique to handle this huge amount of data thereby the researchers can able to do analyze and interpret them. The accuracy of the resultant value perhaps not up to the level of expectation when the dimensions of the dataset is high because we cannot say that the dataset chosen are free from noisy and flawless. So it is required to reduce the dimensionality of the given dataset in order to improve the efficiency and accuracy. Moreover the running time of an algorithm certainly has to be minimized to achieve the desired results. This is being done by apply the same data set to a same clustering technique with and without performed the dimensionality reduction technique principal component analysis on original data. The results of the two approaches are compared and it is proved that the results of clustering using PCA are more accurate, easy to understand and above all the time taken to process the data was substantially reduced.