基于改进神经网络挖掘算法设计.docVIP

  • 3
  • 0
  • 约7.42千字
  • 约 12页
  • 2018-08-29 发布于福建
  • 举报
基于改进神经网络挖掘算法设计

基于改进神经网络挖掘算法设计   摘 要: 针对传统的数据挖掘算法存在结构复杂、耗时长、数据分析过程中易出现错误,数据计算结果难以准确表达结果等缺陷,提出结合神经网络在数据挖掘中的应用方法。由于神经网络拥有对噪声数据承受能力高、错误率低等优点,因此结合神经网络系统对数据挖掘算法进行改进设计可大幅度提高数据准确性,该方法拥有结构简单、表述清晰、精准度高等优势。为基于神经网络的数据挖掘算法的可行性进行了严谨的实验分析,对实验数据进行认真的记录和研究,实验结果表明,基于神经网络的挖掘算法相比传统数据挖掘算法,其精度明显提高,且整个过程耗时较短,由此可证实基于神经网络的数据挖掘算法具有更高的实用性。   关键词: 数据挖掘; 神经网络; 粗糙集; 数据挖掘算法; 数据计算; 可行性分析   中图分类号: TN711?34; TP183 文献标识码: A 文章编号: 1004?373X(2018)14?0143?04   Design of mining algorithm based on improved neural network   HUANG Wenfeng   (Henan Provincial Institute of Scientific Technical Information, Zhengzhou 450003, China)   Abstract: In allusion to the shortcomings existing in the traditional data mining algorithm for its complex structure, long time consumption, error easily appearing in data analysis process, and difficulty in accurate result expression of data calculation results, an application method of combining neural network in data mining is put forward. As the neural network has the advantages of strong ability to withstand noise data, and low error rate, improvement design of the data mining algorithm by combining with the neural network system can greatly improve the data accuracy. This method has the advantages of simple structure, clear expression, high precision and so on. A rigorous experiment analysis for the feasibility of the data mining algorithm based on neural network is performed on the basis of the recorded experimental data. The experimental results show that the mining algorithm based on neural network has higher accuracy and shorter time consumption than the traditional data mining algorithm, which can confirm that the data mining algorithm based on neural network is more practical.   Keywords: data mining; neural network; rough set; data mining algorithm; data calculation; feasibility analysis0 引 言   数据挖掘算法是从数据库的数据中提取隐含有用信息的过程。现阶段常用的数据挖掘算法和理论主要集中于粗糙集理论和遗传算法等相结合的方式[1]。为了解决传统方法结构复杂、耗时长、错误率高等问题,提出改进神经网络的数据挖掘算法。首先对神经网络进行优化和改进,以便减少误差,然后基于改进后的神经网络设计数据挖掘算法。该算法拥有鲁棒性强、精准度高、抗噪能力强、数据承受能力

文档评论(0)

1亿VIP精品文档

相关文档