Empirical Analysis of Genetic Evolution Algorithm Based on Multi-Layer Chromosome of Gene Expression Programming and K-Means Clustering Algorithm

As hypertension has become one of the principal diseases affecting human health, and its prediction accuracy is a topic of concern for the medical professionals, so the computer-aided prediction model of target organ damage of primary hypertension is worthy of research. GEP model with simple chromosome, linear, compact, easily genetic manipulation can eliminate the correlation of gene expression inputs in a variety of training samples. And it has the intelligent, more flexible model structure, having wider methodological applicability, higher prediction accuracy and other characteristics. This study shows that the prediction model has a great prospect for application in the auxiliary prediction of target organ damage in primary hypertension. And classification accuracy is 87.5%, which is higher than SVM and BP neural networks.