A Survey on Analysis of Genetic Diseases Using Machine Learning Techniques

Abstract

The approach of new technological developments in the genetic disease repository has facilitated genetic disease treatment. In the post-genomic time gene detection, which causes genetically excessive diseases, is one of the greatest deterrent tasks. Complex diseases are frequently very heterogeneous and make biological markers difficult to identify. Markers commonly depend on the Machine Learning Algorithms to define, but their success completely depends on the quality and dimensions of the data present. In the machine learning area, computers are promised to support people and analyze large and complex data systems primarily for the production of practically enhanced algorithms. A supervised machine learning methodology has been developed to predict complex genes that cause disease and experiment with the developed algorithm to improve and identify genetic classifications that engage in complex diseases. Genetic Disease Analyzer (GDA) was de veloped using machine learning using the Principal Component Analysis (PCA), Random forest, Naive Bayes and Decision Tree algorithms and the results were compared. The accuracy of 98.79% and sensitivity of 98.67% for the GEO data set is provided for the GDA model. The results of machine learning approaches were examined and their practical applications were discussed in the study of genetic and genomic data.