Quantitative Collection Method Of Dietary Intake And Intelligent Analysis Of Nutritional Ingredients And Chronic Diseases Based On App

With the development of economy,people’s daily diets become more and more abundant,but the abundant diet does not mean health.Unhealthy diets and unbalanced nutrients intake will have a serious impact on people’s health.In recent years,the threat of chronic diseases to human health has become more and more serious and studies have shown that food intake is an important factor leading to chronic diseases.Food intake can be understood as nutritional ingredients intake.Nutritional ingredients are substances in food that provide energy for the survival and growth of the human body and regulate the physiological state of the body.Lack of nutrition or overnutrition can lead to chronic diseases,so how to prevent and treat chronic diseases through nutritional analysis has become the focus of attention.In recent years,the state has actively promoted “Internet + Nutritional Health” to explore a new direction of nutritional health development.The Ministry of Science and Technology of the People’s Republic of China also clearly listed a series of research directions on the quantitative methods of food nutrients and intelligent analysis in the key research and development plans for 2020.In this context,this thesis combines App,machine learning and nutritional ingredients with chronic diseases to carry out the following research:(1)Based on the difficulty,time-consuming,laborious and inaccurate status of traditional nutritional ingredients data acquisition,a nutritional ingredients data acquisition system based on App is proposed in this thesis.Five functional modules of the system are designed and implemented.The database structure of the whole system is designed and implemented through My SQL database and other related technologies.The nutritional ingredients data acquisition system is designed to obtain nutritional ingredients data quickly and accurately,while providing professional nutritional analysis for users involved in data collection.(2)This thesis explores the relationship between nutritional ingredients and chronic diseases.Taking hypertension as the most serious threat to human health in chronic diseases as an example,based on the related technology of machine learning,a five-stage method based on XGBoost for analyzing nutritional ingredients intake to predict hypertension is proposed.In the last stage,this thesis builds a classification model of hypertension based on XGBoost,and compares it with four commonly used classification algorithms of machine learning.The results show that the accuracy rate and F1＿score of hypertension classification model based on XGBoost are the highest,and the classification effect is the best.Compared with other models for predicting hypertension,the results are comparable or even better,which proves that it is feasible to predict and classify hypertension by analyzing human nutritional ingredients intake data.In addition,10 of the 26 nutritional ingredients that mainly affect hypertension are obtained through the characteristic importance analysis of XGBoost.