基于朴素贝叶斯算法的电网建设资源自动化分类模型研究

2021-01-07 13:36:16夏常明
粘接 2021年12期

夏常明

摘 要:提出基于朴素贝叶斯算法的电网建设人力资源自动化分类模型研究。通过网络爬虫技术,采集电网建设资源,在采集的资源信息中提取资源特征,并应用随机森林算法在特征项中选取合适的特征项,生成特征子集。采用朴素贝叶斯算法,构建基于属性相关性度量的分类模型,实现电网建设人力资源自动化分类。结果表明,在两分类和多分类条件下,文中提出的分类模型的平均适应度值分别为89.78和97.47,该分类模型能够获取准確的自动化分类结果,提高模型的适应度值,满足电网建设需求。

关键词:朴素贝叶斯算法;电网建设;自动化分类模型;特征属性

中图分类号:TP301 文献标识码:A     文章编号:1001-5922(2021)12-0093-05

Research on Automatic Classification Model of Power Grid Construction Resources based on Naive Bayes Algorithm

Xia Changming

(State Grid Gansu Electric Power Company, Lanzhou 730030, China)

Abstract:The automatic classification model of power grid construction resources based on naive Bayesian algorithm is proposed. The network crawler technology is used to collect the power grid construction resources, and the resource features are extracted from the collected resource information. In addition, the random forest algorithm is applied to select the appropriate feature items, which are used to generate the feature subset. The naive Bayes algorithm is adopted to construct a classification model based on attribute correlation measurement. Thus the automatic classification of human resources in power grid construction is realize. The results show that under two classification and multiple classification conditions, the average fitness of the proposed classification model is 89.78 and 97.47, respectively. The classification model can obtain accurate automatic classification results, improve the fitness of the model, and meet the requirements of power grid construction.

Key words:Naive Bayes algorithm; Power grid construction; Automatic classification model; Feature attributes

0 引言

考虑到电网建设的复杂性,施工过程中需要大量工作人员进行配合,电网建设资源中人力资源分类管理成为研究重点问题。其中,不同类型的工作人员需要在不同的施工阶段进行有效配合,根据人力资源分类结果,明确人员经验和熟练程度,将其分配至合理岗位,保证电网建设项目的高效开展,并且有利于成本控制[1-2]。

目前相关领域学者针对电网建设资源管理中的分类模型进行了研究,并取得了一定的研究成果。文献[3]利用强化学习技术中的Q学习算法,建立增量分类模型。通过对样本增量序列的有效选取,将分类数据中包含的噪声影响降低,完成高精度的样本自主标记。结合批量增量计算方法,保证分类模型计算复杂度的降低,但该模型的分类精度较低。……

登录APP查看全文