Análisis y programación de la similaridad de Lerman entre variables binarias

Abstract

The Implicative Statistical Analysis started more than 40 years ago, which contains methods such as the so-called Lerman similarity that works with binary variables and was automated in the CHIC computer program; then, a package was created in R (RCHIC), but the Implicative Statistical Analysis (ASI) does not have a programming in the Matlab software to help engineers and mathematicians with the analysis of the large amount of data that is generated nowadays. Therefore, the aim of this research was to program the Lerman similarity between binary variables in Matlab and validate it through the similarity existing with the RCHIC package. The methodology implemented was quantitative and a RGXO1 correlational type, pre-experimental design was used, since the independent variable is not manipulated due to the existence of a programming with which it is compared; the study group consisted of 100,000 random bases of binary data, from which a sample of 383 bases was selected by random sampling and analyzed in the Matlab program. By means of this methodology, it was possible to obtain the Lerman similarity programming in the Matlab software; then, all the binary databases were analyzed by making a comparison between the similarity matrices and the significant nodes. In conclusion, the programming carried out in Matlab obtained a ninety-five percent similarity with the RCHIC package, so it is recommended to improve its programming for future studies.