Research On Elderly Health Knowledge Discovery With Ensemble Learning And Explanation Methods

According to the national plan for the development of ageing undertakings and the construction of the old-age pension system,it is expected that by 2020,the population of the aged over 60 will increase to about 255 million people,accounting for about 17.8% of the total population.Population aging is becoming a very serious social problem,which affects the social and economic development of our country.With the increasing age of the elderly,the health problems of the elderly will become increasingly prominent.The inconvenience of movement and the risk of suffering from various chronic diseases will increase,which will lead to the increase of medical and nursing costs.Therefore,the research on the health needs of the elderly is an important direction for researchers.How to find out the influencing factors of the health status of the elderly,identify the potential factors and high-risk factors,and the interaction mechanism between them,so as to provide a scientific and comprehensive basis for the government to make decisions and formulate targeted policies,so as to better protect the rights and interests of the elderly and improve the quality of life of the elderly,is one of the important issues that researchers need to consider.With the arrival of the data age,people pay more and more attention to how to acquire knowledge from existing data,which has led to the rapid development of artificial intelligence technology represented by data mining.At the same time,data mining has been widely used in various industries,and provides new ideas and methods for solving industry problems.In this paper,data mining method is applied to the analysis of health impact factors of the elderly in China,and the data mining technology is used to analyze the health impact factors of the elderly.Firstly,this paper introduces the research status of the influencing factors of elderly health,and the application of data mining in knowledge discovery.Then,through the analysis of the data from the 2014 China Old Population Health Longevity Survey(CLHLS),the key features are preliminarily acquired with the idea of ensemble learning,an ensemble feature selection method is proposed.The ensemble model is used to analyst data,and different from the general data mining research which only focuses on the efficiency and accuracy,and the interpretability and comprehensibility of the results are relatively less concerned and lack of understanding of the data.We combine the ensemble model with the latest methods of data mining interpretative research and analyze the results in detail from different perspectives.Finally,based on the results of this study,it provides decision-making suggestions for healthy old-age care.The experimental results show that the proposed method can effectively distinguish the elderly with different health status,and identify the various factors affecting the health status of the elderly and their roles.It has important reference significance for knowledge discovery and assistant decision-making of elderly survey data.In addition,this paper uses complex data mining methods to study the data of elderly health survey,which provides a reference paradigm for the application of data mining methods in relevant social survey data and has an important significance.