This bachelor thesis provides an introduction to multivariate statistics, which is the analysis of data with multiple variables using statistical methods. The thesis focuses on the generalization of the normal distribution to random vectors, properties of the multivariate normal distribution, and non-parametric kernel estimation methods for estimating densities. Additionally, the thesis presents methods for separating populations and classifying new observations within these populations using multivariate statistics. The applications of these methods is demonstrated using a real data set, and the accuracy of the classification is evaluated.