时间序列分析在经济投资中的研究与应用

发布时间：2018-08-01 16:53

【摘要】：自从上世纪80年代至今，外商直接投资(FDI)活动日益增加，使其逐渐变成国际资本流动的重要形式。在FDI的带动下，欠发达国家经济技术得到了迅猛的提高。在经济全球化以及我国改革开放策略的影响下，我国成为利用FDI的大国。因此，研究外商直接投资，以此为根据做出合理、有效的决策在新时期具有非常重要的意义。时间序列分析是在数理统计的基础上逐渐发展并成熟的学科，在经济领域应用非常广泛。时间序列分析可以对我国外商直接投资情况进行建模，并对其进行预测，为政府和投资者提供决策依据。本文应用时间序列分析理论对我国外商直接投资情况进行了数学建模，并进行预测。主要内容包括： 1.深入的研究我国FDI数据的变化规律，并针对其建模。通过对实际资料的分析，可以发现我国FDI数据随时间发展的规律，其中针对FDI数据易受噪声干扰的问题，利用小波分析剔除噪声对预测的影响。应用MATLAB和Eviews6.0软件对我国外商直接投资数据进行仿真实验，验证了该方法在计算精确度上优于基本时间序列分析方法。 2.综合了线性和非线性时间序列模型的优势，对我国外商直接投资情况进行了实例分析和建模。由于线性模型往往只描述了自相关性，忽视了异方差性，，而非线性模型可以较好的解决这一问题，因此在线性模型的基础上加入非线性模型可以将FDI数据的变化规律更加科学和全方位的展示出来。 3.在时间序列分析的基础上融入干预分析，使模型更符合实际。在实际生活中，经济数据常常可能受到突发事件的影响，会对时间序列分析造成消极影响，建立干预模型可以消除干预影响。针对经济危机对我国外商直接投资数据的干预影响，应用Eviews6.0软件建立了干预模型，对剔除干预后的数据建立了时间序列模型。通过对预测结果的比较，发现干预分析降低了我国外商直接投资形势的预测误差，提高了预测精度。
[Abstract]:Since the 1980s, the (FDI) activities of foreign direct investment have been increasing day by day, which has gradually become an important form of international capital flow. Driven by FDI, the economy and technology of less developed countries have been rapidly improved. Under the influence of economic globalization and China's reform and opening up strategy, China has become a big country utilizing FDI. Therefore, it is of great significance to study foreign direct investment and make rational and effective decisions in the new period. Time series analysis is a mature discipline based on mathematical statistics, which is widely used in the economic field. Time series analysis can model and predict the situation of foreign direct investment in China, and provide decision basis for the government and investors. In this paper, the time series analysis theory is used to model and predict the situation of foreign direct investment (FDI) in China. The main contents include: 1. The change law of FDI data in our country is deeply studied, and the model is built for it. Through the analysis of the actual data, we can find the law of the development of FDI data over time in our country. Aiming at the problem that FDI data is vulnerable to noise interference, wavelet analysis is used to eliminate the influence of noise on prediction. The MATLAB and Eviews6.0 software are used to simulate the foreign direct investment data in China. The results show that this method is more accurate than the basic time series analysis method. 2. Based on the advantages of linear and nonlinear time series models, the foreign direct investment (FDI) in China is analyzed and modeled. Because the linear model only describes the autocorrelation and ignores the heteroscedasticity, the nonlinear model can solve this problem better. Therefore, the nonlinear model can be added to the linear model to show the variation of FDI data more scientifically and comprehensively. 3. On the basis of time series analysis, intervention analysis is incorporated to make the model more realistic. In real life, economic data may be affected by unexpected events, which will have a negative impact on time series analysis, and intervention model can be used to eliminate the impact of intervention. In view of the impact of economic crisis on FDI data in China, the intervention model is established by using Eviews6.0 software, and the time series model is established for the data after excluding the intervention. Through the comparison of the forecast results, it is found that the intervention analysis reduces the forecast error of the situation of foreign direct investment in China and improves the forecast precision.
【学位授予单位】：沈阳工业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.6;O211.61;F224

【参考文献】