Minimal Learning Machine for Interval-Valued Data

Abstract

Solving regression problems with interval-valued datasets is a challenging task that may arise in many real world applications. Motivated by that fact, many researchers have proposed nonlinear regression methods to handle interval-valued data in recent years. In this paper, we propose two variants of the Minimal Learning Machine (MLM) for interval-valued data. The choice of MLM is explained by its remarkable performance in many applications and the need of a single hyperparameter definition. We present a performance comparison between our methods and five benchmark nonlinear regression methods. The proposed methods presented competitive results.