A Semi-supervised regressor based on model trees

Plenty of Machine Learning (ML) approaches have been applied to a variety of tasks in the past few years, leading to several kinds of algorithms that deal with specific problems efficiently. However, the inability to generalize data with only few instances adequately remains an open problem/ subject. Thus, the interest of data scientists has been shifted towards ensemble learners, which try to leverage the predictive behavior of the different algorithms, offering improved learning performance and robustness. In this study, an ensemble semi-supervised regressor that operates using/with a small number of labeled data is presented. The proposed algorithm is useful when there is a shortage of labeled data, while at the same time offers accurate predictive behavior (with a small time complexity?) without spending much computational time, using the M5 algorithm as a central learner.(1)