Research On Medical Service Cost Forecasting Model Based On BP Neural Network

With the continuous development of medical and health care and improvement of health system reform,medical insurance payment reform is one of the important supporting measures of health system reform,which has great significance to regulate medical service behavior and enhance people's sense of access to health care.Medical cost prediction as a complex system engineering,its influential factors are complex and individual differences are large.This is no doubt that it is difficult to predict accurately.As a result of faster and faster rising medical costs,this issue has aroused the concern of governments.The reason is due to the serious increase in social aging and not comprehensive of medical supervision system.The United States was the first to realize that measures should be taken to control excessive costs and unreasonable medical practices.After years of research and practice,the current international has formed a more mature medical insurance payment system,which the diagnosis level,length of hospital stay,complications,etc.as the core,that is,according to the disease group payment.Compared with the mature international experience,at present,China's medical expenses control,especially in terms of insurance payment has not yet formed a complete methods and technical system,which is in line with China's national conditions.Existing research and technology cannot meet the reality of medical institutions,especially the majority of patient's demands,so actively looking for more scientific prediction method has important theoretical significance.Based on the payment of disease group basis,combined with other factors we create a medical cost forecast model,which is based on the artificial neural network BP algorithm.The ability of neural networks to deal with nonlinear problems is very strong.Using the neural network super self-learning ability,from a large number of patient's actually cost,we can find and grasp the complex objective laws between the cost and its main influencing factors,which provides scientific and rational analysis ideas for China's medical expenses forecast and guidance for patients.Based on the research results at home and abroad,this paper firstly clarifies the research topic of this paper from the aspects of disease group payment theory,laying the foundation for the follow-up study.Secondly,starting from the basic principle of neural network,we demonstrate the feasibility of neural network for cost forecasting,introduce the principle and idea of forecasting medical expenses with neural network,and establish a prediction model based on BP neural network using MATLAB programming technology.For this model,we establish an index system of the corresponding major affecting factors to cost,provide these indicators to quantify the system.Finally,carry out empirical analysis based on the data of the patient's medical records in a certain period of time,establish a prediction model based on BP neural network,and get conclusion after simulation.