《基于BP人工神经网络的企业人员素质综合评价模型研究》由会员分享，可在线阅读，更多相关《基于BP人工神经网络的企业人员素质综合评价模型研究（69页珍藏版）》请在金锄头文库上搜索。
1、 I摘 要 摘 要 人员素质评价是现代人力资源管理的一项重要功能。如果说人力资源管理可以获取组织竞争优势的话，人员素质评价和选拔无疑是实现人力资源管理这种功能的有效途径之一。在人力资源管理的研究和实践中，人员素质指的是那些影响员工工作绩效的自身条件和因素，它是个体完成任务、形成绩效和继续发展的前提。人员素质评价通过对被评价者的素质进行评价分析，为企业的人事配置提供科学依据，辅助企业客观有效地选用人才，为企业的人事决策提供有效参考信息， 对企业进行人力资源的配置、 使用及培训开发有着重要的意义。 人员素质评价在不同程度上受到多种因素的影响，评价结果难以用恰当的数学解析式来描述，是一个多变量、模糊
2、复杂的非线性过程。目前国内外常用的评价方法和技术大多存在评价结果不够客观、准确性差等问题，影响了人员素质评价方法技术在企业人力资源管理实践中的应用。 自上世纪 80 年代以来，人工神经网络这个多学科高科技领域，吸引了众多的神经生理学家、心理学家、数理科学家、计算机与信息科学家及工程师和企业家等进行研究和应用，并在信号与图像处理、语音识别、虚拟现实、控制系统设计、系统仿真、人工智能、优化计算、企业危机管理与预警、数据挖掘、系统辨识及综合评价等领域都得到了很好的应用。 本文尝试利用人工神经网络本身具有并行处理数据、良好的容错、自适应和自学习以及较好的非线性功能等特性，对企业人员素质结构进行系统的分
3、析，构建人工神经网络模型，对企业人员素质评价方法进行改进，以减少由于测评人员主观因素造成的评价结果的偏差，期待能取得一个具有通用性、简洁性的评价企业人员素质的客观量化标准，力图在评价方法上有所创新。 本文在阅读大量前人书籍和文章的基础上，主要做了以下工作： 1概述了企业人员素质评价的常用方法，整理和分析了神经网络的有关理论和相应算法，在分析企业人员素质结构和评价的传统方法的基础上，对企业人员素质评价方法进行了改进研究。 2论证了人工神经网络用于人员素质评价的可行性，介绍了用神经网络对企业人员素质评价方法改进的原理及思路，并以 BP 神经网络模型为基础建立了评价模型；针对人工神经网络评价模型，建
4、立相应的企业人员素质评价的指标体系，并提供将这些指标体系进行量化的方法，使其能够作为神经网络训练数据的理想输入。 3简要说明了在本文研究过程中遇到的问题，阐述了研究的不足之处，提出了继续完善的部分设想，为接下来进行此研究的人提供参考。 关键词：企业人员素质；人工神经网络；素质模型 IIAbstract Personnel competence evaluation is an important function of modern human resource management. In theory and practice of human resource management, p
5、ersonnel competence refers to the conditions and elements that affect the job performance of personnel. It is the base on which the individual can complete the job, form the job performance and develop further. By analyzing the ability of the personnel being evaluated, we can use personnel ability e
6、valuation to offer scientific personnel allocation, aiding the companies selecting personnel objectively and effectively, fully developing the initiative and creativity of personnel, and finally offer reference information for personnel decisions. It is important for company human resource allocatio
7、n, using, and training. Personnel competence evaluation is affected by various factors to some extent, so the outcome of evaluation is difficult to be expressed with a mathematical function; it is a multivariate, complex and nonlinear. Currently, the general fact is we usually get less objective and
8、 correct result by using the evaluation methods and techniques often adopted by researchers home and abroad. So the spread of personnel competence evaluation in company human resource management is greatly limited. Since the 80s of last century, the high techniques multi-subject concept of artificia
9、l neural network has attracted the attention of many neuron-biologists, psychologists and other scientists in mathematics, engineers and entrepreneurs; they are interested in this concept and have conducted researches on it. As the reaches go on, this concept has been widely used in such areas as si
10、gnal and image processing, speech recognition, virtual reality, control system design, system simulation, artificial intelligence, company crisis management and early-warning, data mining, system identification and synthetic evaluation, etc. As artificial neural network possesses such advantages as
11、parallel data processing ,fault-tolerant ,self-adaptive and self-learning and nonlinear, This paper is aimed to analyze personnel ability structure systematically and construct the artificial neural network model, in hope to improve the evaluation process, decrease the possible outcome differences c
12、aused by subjectivity of evaluators and finally invent a objectively quantified standard for evaluating the personnel ability generally and concisely. In preparing this paper, I have read huge volumes of classics and papers, and my work could be summarized as follows: III1. Generally introduced the
13、ordinary methods of personnel ability evaluation, processed and analyzed relative theories and algorithm of artificial neural network, improved the method of personnel ability evaluation based on analyzing personnel ability structure and traditional means for personnel ability evaluation. 2. Analyze
14、d the probability of artificial neural network; introduced the principles and methods by which improving the current personnel evaluation with artificial neural network; constructed the new evaluation model based on BP artificial neural network; with the help of this model, created the index system
15、of personnel ability evaluation, and then proposed means to quantify such indexes, then the indexes can be used as ideal input for artificial neural network training. 3. Mentioned the difficulties during the research and the disadvantages, proposed ideas to further improve the research as the refere
16、nce for the follow up researchers on this subject. Key words: Company personnel ability; Artificial neural network; Competence model 基于 BP 人工神经网络的企业人员素质综合评价模型研究 图 表 索 引 图 表 索 引 图 2.1 BP 网络拓扑结构(11) 图 2.2 S 型曲线 (12) 图2.3 BP网络学习流程图 (15) 图 4.1 企业人员素质人工神经网络评价模型建模流程图 (32) 图 4.2 三层 BP 神经网络参数设置界面(32) 图 4.3 神经网络学习过程算法流程图 (33) 表 4.1 等级评分表 (2
