Decision trees most commonly are univariate, i.e., they use splits based on a single attribute at each internal node. Multivariate decision trees can use splits that contain more than one attribute at each internal node. Though several methods have been developed in the literature for constructing multivariate trees, this body of work is not as well-known as that on univariate trees. We summarize below the directions work on automatically constructing multivariate trees has taken.