Font Size: a A A

Application Of Principal Component Analysis And BP Neural Network Research On Health Expenditure Influenced By HMO Market

Posted on:2011-03-21Degree:MasterType:Thesis
Country:ChinaCandidate:S LiaoFull Text:PDF
GTID:2144360305963635Subject:Industrial Economics
Abstract/Summary:PDF Full Text Request
The rapid increase of health expenses has become a global problem. How can we have a good medical insurance system, but does not cause the financial burden too be overweight? The control of the health expenses is one of most essential factors. U.S. Health Maintenance Organization (HMO) play a important role in health expenses control, so we could learn from it.After a long period of development, HMOs in U.S. have steady form and system, and have complete statistical data that we can refer to. Therefore, the establishment of analysis model using existing data to predict health expenses of United States enables us to analyze the growth trend of medical costs,study the impact of the HMO market structure on health expenses. It provides a new approach for the quantitative analysis of Health Maintenance Organizations. In addition, the prediction of health expenses under the influence of Health Maintenance Organizations will help analyze the development trend of medical costs, control medical cost growth, and is conducive to resolving the problem of excessive growth of medical costs. Therefore our research has remarkable practical significance.In this paper, based on industry economics, we use the traditional analytical framework of the theory of industrial organization, econometrics and mathematical programming analysis tools, learn from domestic and foreign leading edge research, analyze the the effect of HMO on health expenses controlling, using 1995-2007 Datas, utilize principal component analysis and BP neural network model to predict U.S national health expenditure in hospitals. First, we analyze the effect of HMO market structure and other health and economic data on U.S national health expenditure in hospitals. Then we build a PCA-BP model using data from 1995-2007. The prediction result has high prediction precise with relative error of 0.25%.Forecast on the U.S. health expenditure shows that it will grow stablely in recent years as growth rate remains at about 6.2%. In 2011, it will be more than 90 billion U.S. dollars; The analysis of impact of HMO market structure is consistent with former research that HMO penetration, HMO concentration, HMO profit share all have some controlling impact on health expenses.Therefore, in order to control the growth of U.S. health expenditure, this paper proposes the following measures:1. Increase HMO penetration.2. Enhance the level of HMO competition.3. Increase the proportion of for-profit HMO.The research also has some significance to the improvement of our health care system.
Keywords/Search Tags:HMO, health expenditure, BP neural network, prediction
PDF Full Text Request
Related items