Abstract

Mahalanobis-Taguchi System and Its Medical Applications

Author(s): Chao-Ton Su

The Mahalanobis–Taguchi System (MTS) is a diagnostic and forecasting technique for multivariate data. MTS does not involve direct learning from a training set, but it settles a classifier by constructing a continuous measurement scale. The construction of an MTS model would not be influenced by data distribution, which is a good property for solving class imbalance problem. This paper elucidates the significance of MTS and demonstrates its application in several medical and healthcare case studies.