基于决策树的企业销售人员招聘模型的研究与实现

发布时间：2018-06-22 17:48

本文选题：决策树 + 招聘　；参考：《宁夏大学》2017年硕士论文

【摘要】：随着计算机技术的快速发展,人们的生活变得便捷,工作效率也不断提高,各行各业越来越多的数据被存储下来。数据挖掘技术就是从原始的、不完整甚至存在错误的数据中发现数据之间的潜在联系,然后从这些联系中发现知识的过程。数据挖掘是多种学科相结合产生的,这些学科包括数据存储的载体数据库技术,数据统计分析,机器学习,神经网络等。本论文开始说明了人才对于企业生存和不断发展的重要性,然后介绍了以销售为主的企业对于销售人才的大量需求,但是在销售人才招聘的过程中,应聘者自身往往不能在很短的时间内表现出自身的优点和缺点;人事部门由于时间或者能力有限等多种原因对应聘者的判断有可能存在着一定偏差,这给应聘者和企业在招聘的过程中带来了阻碍。与此同时企业存储的大量的销售员基本属性信息得不到有效利用,于是本文决定采用数据挖掘分类技术,利用企业现存的销售员相关数据在应聘者中筛选适合本公司销售岗位的人才。然后介绍了数据挖掘中的分类方法,其中包括决策树分类方法,朴素贝叶斯分类方法以及神经网络分类方法,并详细介绍了ID3算法、CART算法、C4.5算法的原理以及决策树剪枝的几种方法,最后以北京某公司A区销售员的相关数据为基础,对数据进行预处理,运用R语言将CART算法和C4.5算法应用到实验数据中生成决策树模型,并做出相关的分析。其中在CART算法中,根据节点的复杂程度对决策树进行剪枝,为的是优化模型的同时尽量避免数据的过拟合现象的出现。最后比较了CART算法模型和C4.5算法模型的准确度,利用C4.5算法生成的模型,实现了应聘销售员筛选系统,为销售企业销售人员的招聘提供理论参考。本文最后对该研究做了总结与展望,说明了该模型存在的一些不足,以及改进的方法。
[Abstract]:With the rapid development of computer technology, people's life becomes more convenient, and their working efficiency is improved. More and more data are stored in all walks of life. Data mining technology is the process of discovering the potential relationship between data from raw incomplete or even wrong data and then discovering knowledge from these connections. Data mining is a combination of a variety of disciplines, including data storage carrier database technology, data statistical analysis, machine learning, neural networks and so on. This paper begins with the importance of talent for the survival and continuous development of enterprises, and then introduces the large demand for sales talents in sales oriented enterprises, but in the process of recruiting sales talents, Candidates themselves are often unable to show their own advantages and disadvantages in a very short period of time. Personnel departments may have certain deviations in their judgment of candidates due to various reasons, such as time or limited ability. This creates obstacles for candidates and companies in the recruitment process. At the same time, a large number of basic attributes of salespeople stored in enterprises can not be used effectively, so this paper decides to adopt data mining and classification technology. Use the existing salesperson data in the company to select suitable candidates for the company's sales position. Then the classification methods in data mining are introduced, including decision tree classification, naive Bayes classification and neural network classification. The principle of C4.5 algorithm of ID3 algorithm and the pruning method of decision tree are introduced in detail. Finally, based on the relevant data of salesmen in A area of a company in Beijing, the data are preprocessed. The cart algorithm and C4.5 algorithm are applied to the experimental data to generate the decision tree model, and the related analysis is made. In cart algorithm, the decision tree is pruned according to the complexity of nodes in order to optimize the model and avoid the phenomenon of data over-fitting. Finally, the accuracy of cart algorithm model and C4.5 algorithm model is compared. The model generated by C4.5 algorithm is used to realize the candidate salesperson selection system, which provides a theoretical reference for the recruitment of sales personnel in sales enterprises. In the end, the paper summarizes and prospects the research, and explains some shortcomings of the model and the improved method.
【学位授予单位】：宁夏大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.13

【参考文献】