Multivariate normal sampling; an approach through principal component analysis

In this paper we discuss the use of Principal Component Analysis (PCA) for simulating random samples from multivariate normal distribution, using mean vector and covariance matrix. Sampling is an important aspect in the field of Statistics. We can generate random samples from various univariate distributions either discrete or continuous. We can also generate samples from bivariate distributions for that purpose there are different tables available. But, sampling in that manner is troublesome. In this article we use PCA; a multivariate technique for the purpose of sampling. Furthermore, various properties related to the multivariate normal data can be verified by simulating the samples.