DIMENSIONALITY REDUCTION FOR BIG DATA AN APPROACH WITH PRINCIPAL COMPONENT ANALYSIS

Abstract

Principal component analysis (PCA) technique is widely used technique for dimensionality reduction in
data analysis. There are many benefits to reduce the dimensions of a dataset in different perspective, like visualization of
data is restricted to 2 or 3 dimensions. Reducing the dimensions of data can sometimes significantly reduce the time
complexity of some numerical algorithms. Besides, most of the statistical models have disadvantage from high
correlation between covariates, PCA have the application to produce linear combinations of the covariates that are
uncorrelated between each other. Principal component analysis (PCA) uses the concept of orthogonal transformation,
which transforms correlated variables of set of observation’s to a result set consisting of values which are linearly
uncorrelated variables called principal components. The result set we get after this transformation are known as
uncorrelated orthogonal basis set.