Fuzzy measure for similarity of numerical vectors

In this paper we describe a method for Fuzzy Measure for Similarity of Numerical Vectors (FMSNV) in order to compare and measure behavior of numerical vectors based on fuzzy methodology. Each numerical vector is treated as a fuzzy set and the method described here is designed to find similarity in behavior of two or more vectors. The method is capable of comparing not only pairs of vectors, but also to compare behavior of one variable (vector) to a group of explanatory variables (vectors), similarly to the modeling performed with the help of econometric methods. In those cases where the number of explanatory variables exceeds one, FMSNV also computes relative importance of each of the explanatory variables used in the analysis of the behavior of variable which is modeled. First, the FMSNV and its underlying methodology and conceptual scope are presented. Second, a case study is presented to demonstrate practical capabilities of the method; the same modeling is performed using traditional econometric method (OLS) and the results are compared to the results generated by the FMSNV to demonstrate the advantages of FMSNV.