GEOGRAPHICAL RESEARCH ›› 2014, Vol. 33 ›› Issue (3): 451-466.doi: 10.11821/dlyj201403005

Previous Articles     Next Articles

Geographic distribution of reference value of boys’ peak expiratory flow rate based on the artificial neural networks

GE Miao, XUE Ranyin, HE Jinwei, HU Yanyu   

  1. Health Geography Institute, College of Tourism and Environment, Shaanxi Normal University, Xi'an 710119, China
  • Received:2013-01-15 Revised:2013-10-24 Online:2014-03-10 Published:2014-03-10

Abstract: To improve the situation in which geographical factors are ignored when healthy boys PEFR values are estimated, this article aims to analyze the relationship between their PEFR reference values and geographical factors. Correlation analysis was adopted in the process of collecting Chinese healthy boys' PEFR values to explore the data and the selected 25 geographical factors. After that, those 10 geographical factors which have correlation with the data were extracted for further analysis. Furthermore, spatial autocorrelation (Moran's index) shows that the data is correlated with spatial and geographic factors. The artificial neural networks were created to analyze the simulation of the selected indicators of geographical elements. This research chooses 5 layer neural networks and selects 9 hidden layers and 1000 times of training to build a simulation rule, and this rule thereafter was used to simulate the relationship between healthy boys' PEFR reference values and geographical environment. The distribution map of reference values was generated by using Arcgis' statistical analysis to test the data's distribution and choosing the disjunctive kriging interpolating. It is indicated that the artificial neural network and geostatistical analyst can be combined to generate a better interpolation map and that the Chinese boys' PEFR values have some correlation with longitude, altitude, annual average temperature, annual average relative humidity, wind speed, average annual soil gravel content, soil organic matter content, soil cation exchange capacity (clay), soil cation exchange capacity (silt), and soil total exchangeable amount. Meanwhile, this article analyzes the relationship of the geographical factors and the medical indicators and discusses the effect of these factors on Chinese healthy boys' PEFR values.

Key words: PEFR reference value, BP neural network, geostatistical analyst, Moran’s index, spatial analysis