人工神经网络在呼吸系统疾病急诊就诊人数预报中的应用

发布时间：2018-06-06 20:36

本文选题：呼吸系统 + 气象因子　；参考：《兰州大学学报(自然科学版)》2014年01期

【摘要】：利用北京市2009-2011年呼吸系统疾病急诊就诊人数资料和同期的气象资料及污染资料,分析了气象因素及污染物分别与上感、下感急诊就诊人数的相关性,在此基础上,通过BP人工神经网络分别建立了上感和下感急诊就诊人数的预报模型,并对其效果进行评价.结果表明:气象因素和污染物与上感、下感的发病有密切的关系;建立的上感、下感就诊急诊人数的神经网络预报模型结构分别为13-7-1(即有13个输入、7个隐含节点和1个输出)和13-6-1(即有13个输入、6个隐含节点和1个输出),预测准确率分别为77.11%和75.57%.与统计预报方法相比较,该方法计算简便、误差较小、预测准确率高,对上感和下感急诊人数有较好的预测效果,为医疗气象预报提供了一种新方法,具有进一步的研究价值.
[Abstract]:Based on the data of emergency attendance of respiratory diseases in Beijing from 2009 to 2011, meteorological data and pollution data of the same period, the correlation of meteorological factors and pollutants with the number of emergency patients with upper and lower sensations were analyzed. Based on BP artificial neural network, the prediction models of the number of emergency patients with upper and lower sensations were established, and the effects of the models were evaluated. The results show that meteorological factors and pollutants are closely related to the incidence of upper and lower sensations. The structure of neural network prediction model for emergency patients was 13-7-1 (13 inputs, 7 hidden nodes and 1 output) and 13-6-1 (13 inputs, 6 hidden nodes and 1 output). The prediction accuracy was 77.11% and 75.57%, respectively. Compared with the statistical forecasting method, this method has the advantages of simple calculation, small error and high prediction accuracy. It has a better prediction effect on the number of emergency patients with upper and lower senses. It provides a new method for medical weather forecast and has further research value.
【作者单位】：兰州大学大气科学学院甘肃省干旱气候变化与减灾重点实验室;
【基金】：中央高校基本科研业务费专项资金(自由探索)项目(LZUJBKY-2012-123) 国家自然科学基金项目(41075103) 公益性行业(气象)科研专项项目(GYHY201106034)
【分类号】：R56

【参考文献】