Optimal predicting method of peritoneal fluid absorption rate using genetic algorithm embedded in neural network

This paper addresses the predicting problem of peritoneal fluid absorption rate in the peritoneal dialysis treatment process of renal failure. An innovative predicting model was developed in this paper, which employed genetic algorithm embedded in neural network to predict the important PFAR index in the peritoneal dialysis treatment process of renal failure. The significance of PFAR and the complexity of peritoneal process are analyzed. Genetic algorithm is used to initial weight and bias of neural network, and then optimal predicting model of PFAR was built based on neural network. This method utilizes the global search capability of genetic algorithm and local search advantage of neural network completely. To show the validity of the model, the optimal predicting model is compared with conventional artificial neural network and multivariate regression method. The simulation results show that the predicting accuracy of the optimal neural network is greatly improved and learning process needs less time.