光纤通信网络非平稳数据智能挖掘仿真研究  

Simulation Research on Non-stationary Data Intelligent Mining of Optical Fiber Communication Network

作　　者：赵建华 ZHAO Jian-hua(Network Center,Changchun Normal University,Changchun Jilin 130032,China)

机构地区：[1]长春师范大学网络中心,吉林长春130032

出　　处：《计算机仿真》2020年第3期324-327,共4页Computer Simulation

基　　金：吉林省科技厅项目(20150307032GX);吉林省教育科学(GH180995)。

摘　　要：探究一种有效的非平稳数据智能挖掘方法,可以克服智能挖掘过程中受多种噪声的干扰,对光纤通信网络的大规模使用具备重要的现实意义。为了解决当前挖掘方法由于各种原因影响造成数据不完整、延时较长、遗漏率较高等问题,提出一种基于关联规则映射的非平稳数据智能挖掘方法,利用原始去噪算法对收集的非平稳数据样本进行去噪,获取非平稳数据的置信度。利用时间加权方法依据非平稳数据置信度对非平稳数据进行去噪,得到去噪后的非平稳数据。将其用于构建数据子空间矩阵,挖掘不同子空间非平稳数据集,利用同一空间下非平稳数据集的关联强度挖掘出非平稳数据集,实现了非平稳数据智能挖掘。仿真测试结果证明,所提方法能够有效缩短挖掘延时、提高数据挖掘完整率、降低遗漏率,具备较强的可行性。An effective method to mine non-stationary data intelligently can overcome the interference of many kinds of noises in the intelligent mining process.It is very important for the large-scale use of optical fiber communication network.In this paper,an intelligent mining method for non-stationary data based on association rule mapping was proposed.Firstly,the original denoising algorithm was used to remove the noise in collected non-stationary data sample,so as to obtain confidence of non-stationary data.According to non-stationary data confidence,time-weighted method was used to remove the noise of non-stationary data,so that the non-stationary data after the noise reduction was obtained,which was used to construct data subspace matrix.In addition,non-stationary data sets in different sub-spaces were mined.Finally,the correlation strength of non-stationary data sets in the same space was used to mine non-stationary data sets.Thus,the intelligent mining for non-stationary data was achieved.Simulation results prove that the proposed method can effectively shorten the mining delay,improve the integrity rate of data mining and reduce the missing probability.This method has strong feasibility.

关 键 词：非平稳数据 智能挖掘 掘延时 完整率 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]