基于人工神经网络的结核病疫情预测研究及软件实现

项目名称： 基于人工神经网络的结核病疫情预测研究及软件实现

项目编号： No.30872160

项目类型： 面上项目

立项/批准年度： 2009

项目学科： 金属学与金属工艺

项目作者： 易静

作者单位： 重庆医科大学

项目金额： 31万元

中文摘要： 本研究利用我国四次全国结核病流行病学调查资料和疫情资料，采用具有高度非线性识别能力的人工神经网络实现方法，探讨建立智能疫情预测模型；并通过不同的计算方法提高神经网络模型预报精度；对SPSS Clementine 进行二次开发，建立Web 版结核病疫情预测系统，实现远程、在线数据挖掘的流程，为政府、卫生行政部门提供强大准确的决策支持，有助于其在制定结核病控制规划中提出适宜的目标和指标值；并实现各结核病防治相关机构之间的资源共享，有利于疫情爆发时的快速反应；为正在开展的结核病控制项目提供科学依据，以期最终实现WHO 提出的"千年发展目标"。

中文关键词： 结核病；神经网络；预测；web 版预测系统

英文摘要： To explore the establishment of intelligent model of predicting tuberculosis prevalence based on four times of national epidemiological investigation results as well as prevalence status of tuberculosis in Chongqing, employing Artificial Neural Network with powerful no-linear recognition ability; to elevate prediction accuracy of Artificial Neural Network through different computing methods; to establish Web-edition prediction system of tuberculosis prevalence through redevelopment of SPSS Clementine in order to carry out long-range and on-line data mining, providing powerful and precise decision support for government and medical administration, helping them to set up appropriate goal and index value in tuberculosis control scheme, implementing resource sharing among predictive and curing agencies of tuberculosis for fast reaction in case of tuberculosis outbreak; to provide scientific foundation for tuberculosis control project undertaken right now in China, in order to materialize "Millennium Development Goal" advocated by World Health Organization in ultimate.

英文关键词： Tuberculosis; Artificial Neural Network; predict; Web-edition prediction system