Pearson’s Correlation

This chapter discusses the essential concepts involved in computing the Pearson’s correlation. Properties of variance, covariance and correlation are discussed first. It is proven that the Pearson’s correlation is \(\pm 1\) when a linear association exists among the variables. A brief discussion of covariance and correlation matrices and conditional correlation follows it. The concept of co-skewness and co-kurtosis is introduced and extended to fractional powers. The chapter ends with application of correlation in DSP and image processing.