...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Crop Research >Back propagation neural network : An interactive tool for effective rainfall prediction
【24h】

Back propagation neural network : An interactive tool for effective rainfall prediction

机译：反向传播神经网络：有效降雨预测的交互式工具

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Natural water source is one of the influencing factors for crop cultivation and production in agricultural sector. At the same time most of the agro industrial development interdepends the natural resources and its impact for its progress. Hence, there exists an ever growing demand for effective rainfall prediction system. The ongoing demand for precise forecast methods has led to thedevelopment of computerized rainfall forecastsscenarios. This investigation makes an attempt to achieve precise rainfall predictions using Neural Network (NN) approach. This researchintroduces a novel exhaustive search based Maximum Frequency Weighted Feature Selection (MFWFS) using approach to identify the significant weather parameterfor prediction.The effects of feature selectionin model performance were also investigated. This was done by examining the performance of the neural networksprediction model using both complete and reduced parameters. A meticulous comparison of the overall performance indicated that theback propagation algorithm approach based neural network model outperformed better than existing methods. The proposed NN architecture achieved 0.0499 error rate and 95.01% prediction accuracy using effective weather parameters. This investigation introduces acompact interactive graphical user interface (GUI) based tool developed using C# and net platform to enable users to conduct meteorological assessment on their own ease. This tool enables users to train and test the NN model for various input options and to visualize the resultsby processing the rainfall forecast scenarios.
机译：天然水源是影响农业部门作物种植和生产的因素之一。同时，大多数农业工业发展都依赖自然资源及其对其发展的影响。因此，对有效的降雨预测系统的需求不断增长。对精确预报方法的持续需求导致了计算机化降雨预报方案的发展。这项研究尝试使用神经网络（NN）方法实现精确的降雨预测。本研究采用一种新颖的基于穷举搜索的最大频率加权特征选择（MFWFS）方法，该方法可以识别重要的天气预报参数进行预测。还研究了特征选择对模型性能的影响。这是通过使用完整参数和精简参数来检查神经网络预测模型的性能来完成的。对整体性能的仔细比较表明，基于反向传播算法的神经网络模型的性能优于现有方法。利用有效的天气参数，提出的神经网络架构实现了0.0499的错误率和95.01％的预测精度。这项调查介绍了使用C＃和网络平台开发的基于紧凑型交互式图形用户界面（GUI）的工具，使用户可以轻松进行气象评估。该工具使用户能够针对各种输入选项训练和测试NN模型，并通过处理降雨预报方案来可视化结果。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号