淮安市水源水质与胃肠道疾病发病关系的 探讨分析

Analysis of water quality relationship with the incidence of gastrointestinal disease in Huai’an

中文关键词:  胃肠道疾病  NARX神经网络  水源水质  拟合

英文关键词:gastrointestinal diseases  nonlinear autoregressive neural network with external input  water quality  fitting

中文摘要:

      目的 探讨水源水质与胃肠道疾病之间的关系, 以及胃肠道疾病的发病规律。方法 选择淮安市2006年1月至2013年12月间水源水质指标及胃肠道患病例数, 并进行相关性分析, 然后进行具有输出时延反馈的非线性自回归(NARX)神经网络模型的构建、训练及模拟。结果 水源水质的石油类、挥发酚、总砷、总汞、6 价铬、氟化物及粪大肠菌群等指标与发病人数有显著的相关性。经过训练后NARX神经网络的可决系数为0.716, 拟合效果比较理想。结论 水源水质与胃肠道疾病之间存在一定的相关性, 基于NARX神经网络对水源水质与胃肠道疾病进行拟合是可行的。

英文摘要:

      Objective To discuss the relationship between water quality and gastrointestinal diseases, and variation of gastrointestinal diseases. Methods Data of water quality indicators and gastrointestinal cases in Huai’an from January 2006 to December 2013 were collected. Correlation analysis was made and the NARX model was built, trained and simulated. Results Petroleum, phenol, total arsenic, total mercury, hexavalent chromium, fluoride and fecal coliform had a significant correlation with the gastrointestinal diseases. The coefficient of determination after the NARX neural network training was 0.716, and the fitting effect was perfect. Conclusion There was a certain correlation between water quality and gastrointestinal diseases. Fitting of water quality and gastrointestinal diseases based on nonlinear autoregressive neural network with external input was feasible.