Two new 3D distance measures for IFSs and their applications in pattern classification and pathological diagnosis

Abstract

This research investigates some prevailing distance measures and discusses their limitations. The main focus of this study is to overcome limitations like the ‘zero divisor problem’ and the counter-intuitive and unreasonable results of the existing distance measures. Two new 3D distance measures for IFSs based on chi-square distance and Canberra distance are proposed to counter such issues. Some set-theoretic features of the proposed measures are discussed. The efficiency of these measures is demonstrated by comparing them with some prevailing distance measures. Furthermore, proposed measures are used for pattern classification problems and pathological diagnoses. The findings show that the proposed distance measures outperform the current distance measures in every aspect.