Research On Building Personal Healthcare Service Platform Using Disease Prediction Model Based On Crowd Cohort Algorithm

To alleviate the pressure of resident’s health and national medical expenditure caused by ageing of population,China has implemented the strategy of "healthy China",aiming to control the disease in the stage of prevention and early treatment,and personal health management platform came into being.Based on the research of platform architecture and disease diagnosis prediction algorithm,the infrastructure of personal health management platform is designed,to solve the problem of low accuracy of the model caused by elderly human sign data deviate from normal value,a high-precision disease diagnosis and prediction model for the elderly was proposed,The main work is as follows:First of all,the development of personal health management and service demand are analyzed,Using the existing technology to design functional modules of personal health management platform,establish the basic framework and prove the advantages of the framework through comparison,then,using the crowd queue algorithm to solve the problems of the algorithm,improve APC(age-period-cohort)model,Based on the results of traditional APC model,the influence of personal habits on the outcome are separated,and personal habits are added to the influencing factors of new APC model,Adopt corresponding parameters for people with different habits.It increases the sensitivity of the model to the partial population(people with certain habits)in the large-scale population cohort analysis.The experiment shows that the improved APC model has higher accuracy than the traditional APC model.Thirdly,the model of disease prediction and diagnosis is established,Taking the quantitative results of the improved APC model as the external impact of social development on the population,GA-clustering(genetic algorithm)was carried out together with the standardized physical data representing the internal changes of individuals Clustering),find out the most similar population by using the crowd queue,complete the disease prediction through the disease change trend of similar population,and prove the advantages of the model in the form of ROC curve.Finally,through the real data of the elderly,taking Cerebrovascular disease and Cardiovascular disease as an example,the experiment and result analysis of the designed model are carried out to prove the feasibility of the model and achieve the design effect.