女性实际碳酸氢根参考值与地理环境的人工神经网络

J4 ›› 2012, Vol. 38 ›› Issue (2): 364-367.

女性实际碳酸氢根参考值与地理环境的人工神经网络

曹晓仪¹|葛淼¹|林天应¹|付成程²

1.陕西师范大学旅游与环境学院地理系|陕西西安 710062；2.陕西师范大学食品工程与营养科学学院食品科学系|陕西西安 710062

收稿日期:2011-09-14 出版日期:2012-03-28 发布日期:2012-03-28
通讯作者: 葛淼(Tel:029-85310532,E-mail:gemiao@snnu.edu.cn) E-mail:gemiao@snnu.edu.cn
作者简介:曹晓仪（1986-）|男|重庆市人|在读理学硕士|主要从事医学正常参考值与地理环境关系的研究。
基金资助:
国家自然科学基金资助课题(40971060)

Artificial neural networks between actual bicarbonate reference value of Chinese women and geographical environment

CAO Xiao-yi¹|GE Miao¹|LIN Tian-ying¹|FU Cheng-cheng²

1. |Department of Geography,School of Tourism and Environment,Shaanxi Normal University,Xi’an 710062,China;2. Department of Food Science,School of Food Engineering and Nutritional Science,Shaanxi Normal University,Xi’an 710062,China

Received:2011-09-14 Online:2012-03-28 Published:2012-03-28

摘要/Abstract

摘要：

目的：探讨中国女性实际碳酸氢根参考值与地理因素间的关系，为制定女性实际碳酸氢根参考值标准和临床诊断提供科学依据。方法：收集中国各地测定的8 138例健康女性实际碳酸氢根参考值，运用相关分析的方法分析其与海拔高度、年日照时数、年平均气温、年平均相对湿度、年降水量及气温年较差6项地理指标的关系，并将地理指标作为神经元输入逆向传播（BP）神经网络，建立复杂非线性模型。结果：上述6项地理指标与中国女性实际碳酸氢根考值有相关性，差异有统计学意义（P<0.05）。通过5层神经网络，经过3 000次训练建立训练模型，用该模型估算出全国4 383个点的模拟值，运用GIS内插绘制出女性实际碳酸氢根参考值分布图。结论：将某地6项地理指标带入模型计算，可获得该地女性实际碳酸氢根参考值。分布图揭示出女性实际碳酸氢根参考值东部高、西部低的地理分异特征。

关键词: 实际碳酸氢根；参考值；地理环境；神经网络；女性

Abstract:

Abstract:Objective To explore the relationship between the actual bicarbonate reference value of Chinese women and geographical environment,and provide scientific basis for the distribution law of actual bicarbonate of women reference value and clinical diagnosis.Methods 8 138 cases of actual bicarbonate reference values were collected.The relationship between reference values of actual bicarbonate of Chinese women and altitude,annual sunshine hours,annual mean air temperature,annual mean relative humidity,annual precipitation,annual range of air temperature were analyzed by correlation analysis.And the geographical factors were as nerve cell input back propagation(BP) neural networks to build nonlinear model.Results The correlation between the reference values of actual bicarbonate of Chinese women and 6 geographical factors was found,and there were significant differences(P<0.05).The reference value of actual bicarbonate of women in 4 383 points all over China were calculated by a model which was established by 5-layer neural network and 3000 time’s training.The distribution map of the reference values of actual bicarbonate was drawn by spatial interpolation.Conclusion The reference value of actual bicarbonate of women can be got by artificial neural networks (ANNs) model.The distribution map reflects that the spatial distribution law of actual bicarbonate reference value in western is lower than that in eastern of China.

Key words: actual bicarbonate;reference value;geographical environment;artificial neural networks

中图分类号:

R188

曹晓仪|葛淼|林天应|付成程. 女性实际碳酸氢根参考值与地理环境的人工神经网络[J]. J4, 2012, 38(2): 364-367.

CAO Xiao-yi|GE Miao|LIN Tian-ying|FU Cheng-cheng. Artificial neural networks between actual bicarbonate reference value of Chinese women and geographical environment[J]. J4, 2012, 38(2): 364-367.