In this section, we explain the computational aspects of the node-splitting procedure, especially for nominal features. Additionally we illustrate how to deal with missing values.