ABSTRACT
There is no reason to limit the denition of a sample space to a single dimension. There are random phenomena that are best described in terms of a vector-valued outcome. In terms of engineering applications, it is easy to imagine that (1) the quality of a manufactured part would be measured by several dimensions such as thickness, height, and length; (2) the demand for consumer electronic products from an inventory would include several types of devices or several distinct models; or (3) the mix of patient needs in a hospital emergency room would imply consumption of various quantities of several resources. In a possibly more familiar application, the cards obtained in a poker hand might also be well described in multivariate terms. Thus, in each of the upcoming example cases, the measured quantities might be best represented as a random vector.
