基于人工神经网络方法的上市公司股票投资价值分析-金融学专业论文.docx

Classified Index: F830.91 U.D.C: 336 Dissertation for the Master Degree in Economics THE ANALYSIS OF INVESTMENT VALUE OF LISTED COMPANIES BASED ON ARITIFICIAL NEURAL NETWORK Candidate： Huiping Zheng Supervisor： Prof. Jin Feng Academic Degree Applied for： Master of Economics Speciality： Finance Affiliation： School of Management Date of Defence： June, 2011 Degree-Conferring-Institution： Harbin Institute of Technology 摘 摘 要 -I- -I- 摘 要 随着中国金融改革的全面推进，股票投资逐渐成为现代社会投资的重要手段， 为我国企业发展募集大量资金，促进社会资源合理配置，对社会经济生活产生重 大而深远的影响。在经济全球化、金融自由化的背景下，我国证券市场朝着规范 运作的方向发展。正确认识证券市场上影响股票投资价值的因素，可以为对上市 公司股票投资价值进行客观及时的评估提供可靠的科学依据，利于投资者有效降 低投资风险，也利于政府对企业的监管和激励，提高市场效率。 本文总结了国内外上市公司股票投资价值评估方法的研究现状和相关理论， 从宏观、中观、微观三个层面对影响股票内在价值的因素进行全面、系统的分析 和归纳，并在此基础上选取了 11 个评价指标作为模型的指标体系。结合 BP 神经 网络相关理论知识，选取我国沪深两市制造业 113 家上市企业进行实证分析。用 自由现金流贴现模型中的定增长模型估计股票内在价值，并对数据进行主成分处 理，减少输入变量，再用试凑法经过两步检验，确定神经网络模型拟合效果最好 的训练函数和隐层节点数，建立较为完善的 BP 网络模型，并使用该模型进行预 测和综合分析。最后根据实证结果，结合中国市场实际情况对模型进行评价，并 根据模型实证研究结果，给出相关政策建议。 关键词：股票内在价值；价值投资；神经网络；BP 神经网络模型 Abs Abstract -II- -II- Abstract With the comprehensive propulsion of Chinese financial reform, stock has became an important investment means of society investment, which raises a large number of capital for Chinese enterprise, promotes reasonable allocation of social resource, makes a great and profound influence on economic life. On the background of economic globalization and financial liberalization, Chinese stock market improves itself toward standard operation. Knowing the influence factors of stock investment value correctly could provide listed company’s stock investment value with reliable and scientific basis, which benefits not only investor to decrease investment risks but also government to supervise and inspire enterprise. This thesis summarize research status and relating theories of domestic and foreign listed company’s stock investment value evaluation method, from three aspects, macro, meso and micro, systematically analyzes and summarizes the influence factors of stock i