A Classification Approach Based on Genetic-Data-Structuring for the Prediction of Hypertension

Hypertension is known to be a major cause of death around the world. The death rate for this disease has increased up to 94% since last decade. Due to this disease, dangerous health conditions arise like heart failure, kidney failure and stroke. To prevent permanent loss, there is an imperious need for automated techniques to be developed for the detection of this disease. Recently, genetic information has been combined with machine learning techniques for the detection of hypertension disease. The genetics based diagnosis also plays a key role in evolution of this disease. In this paper both the genetic and demographic data is used forth detection of this disease. For detection purpose we have gathered genetic datasets of 250 patients from different databases (online repositories). Our research makes use of a feature set comprising genes, SNP alleles, risk alleles, chromosomes, region and nationality of patients. Here, an algorithm is proposed for data structuring. For performance evaluation of our proposed approach we have applied different types of classification models; naive bayes, filtered classifier, LWL, K* and NBF network. The results obtained show that best accuracy is achieved using naive bayes i.e. 98%.