Journal of Statistical and Econometric Methods

A Standardized Treatment of Binary Similarity Measures with an Introduction to k-Vector Percentage Normalized Similarity

  • Pdf Icon [ Download ]
  • Times downloaded: 147
  • Abstract

    This paper attempts to codify a standard nomenclature for similarity measures based on recent literature and to advance the field of similarity measures through the introduction of non-binary similarity between more than two attribute vectors. The nomenclature standardization is accomplished through the integration of common terminology into non-binary similarity measures, and the refinement of the terminology with regard to k-vector binary and non-binary measures. This nomenclature standardization lays the groundwork for the introduction of k-vector percentage normalized similarity measures that follow the same fundamental form as pre-existing binary measures; a method not previously documented.

    Mathematics Subject Classification: 68; 91
    Keywords: Binary Similarity; Nonbinary Similarity; Nonparametric Similarity Testing; Multivector Similarity