Research And Implementation Of "Three-high" Risk Assessment Based On Machine Learning

Health data has a high value. Using machine learning technology to extract information from health data can bring new ideas for disease prediction, disease risk assessment and disease diagnosis. In a sense, the application of machine learning in health data has an important role in improving human living standards.Hypertension, hyperglycemia and hyperlipidemia and its complic?ations have caused serious harm to people’s health. The prevalence rate of"three-high" in our country is increasing year by year. At present, it has become an important public health problem in our country. It’s necessary to reduce the harm through prevention. However, As the "three-high"pathogenesis is complex and difficult to predict through the clinical, it is necessary to use machine learning technology to get the valuable information in health data. By studying the "three-high" prediction model and complete the "three-high" risk assessment, early prevention and treatment can be achieved. In short, the machine learning technology provides a scientific basis for risk assessment of "three-high" and this is exactly the purpose of this paper.The research work of this paper mainly focuses on the design of three-high risk assessment model, the choice of algorithm, the improvement of model and the evaluation of performance. First, a"three-high" risk assessment model was proposed based on the Harvard cancer risk index and logistic regression model.Then, by analysis of the"three-high" obesity-related data, the relationship between obesity-related body measurement parameters and the "three-highs" is proposed. Finally,according to the results of data analysis, combined with decision tree and logistic regression algorithm, another "three-high" risk assessment model is proposed, which makes up the shortcomings of the previous model. By comparing the performance of the two models, the feasibility and performance of the latter model are verified.Finally, the feasibility of the two risk assessment models and the improvement of the performance of the algorithm are verified by experiments. Compared with the former model, the accuracy, specificity and sensitivity of the latter model were improved, especially the improvement of female hyperglycemia risk assessment model. The research results of this subject have a certain guiding significance for the risk assessment of "three-high".