基于人工神经网络的大学生体脂百分比预测方程的构建

Predictive Equation for Body Fat Percentage in College Students Based on Artificial Neural Network

摘要: 以1 201名大学生为研究对象，采用人工神经网络（ANN）数据挖掘方法，构建不同性别大学生体脂百分比（BFP）的预测方程，探讨身体质量指数（BMI）与BFP之间是否为线性关系。结果显示：ANN构建的基于BMI预测BFP方程，精度都高于91%；无论男生还是女生，构建的二次方程拟合值精度高于线性方程。构建的方程为中国大学生营养评价、身体形态健康管理以及相关疾病的危险预测提供依据。

Abstract: The aim of this study is to establish the predictive equation for body fat percentage (BFP) among university and college students, trying to analyze the relationship between body mass index (BMI) and BFP by using artificial neural networks (ANN) and big dataset of 1201 contemporary college students of China.The results showed that BFP productional functions structured by ANN with BMI have the accuracy of higher than 91%.The fitting precision of quadratic equation is higher than that of linear equation in both genders.The established equation provides a more scientific basis for nutritional assessment, health management and risk prediction of related diseases of contemporary college students in China.