Machine Learning Based Genetic Decision Making Methodology Using Genotype-Phenotype Mapping

Abstract

A methodology titled "Dynamic Selection of Essential Similar Principal Components" is presented in this study. This methodology evaluates the similarities while omitting the differences among features to accommodate for all possible similarities caused by genes. It has been tested on real data set collected from the dysmorphic facial images published in scholarly journals, thus accounting decent diagnostic information about the syndrome. The methodology has been tested with 15 different syndromes that accommodate 5 examples per syndrome. It can be concluded based on the results that a great number of syndromes indicating a characteristic pattern of facial anomalies can be typically diagnosed by employing the approach we propose in this study.