Linear regression of interval-valued data based on dp metric

This paper proposes a new interval-valued linear regression model (IVM) based on dp metric. The model treats interval-valued variables as a whole, representing the relationship between one or more explanatory variables and the response variable using a single equation. The use of dummy variables in the estimation of the parameters has two advantages, it addresses the issue that model performance may be affected by the scalar multiplication of interval-valued data, and it ensures the mathematical consistency of the model predictions. Experiments with simulated and real interval data sets demonstrate the merits of IVM.