应用X-12-ARIMA与SARIMA模型及其组合模型对中国保费收入的预测研究
发布时间：2018-12-12 21:24
【摘要】：随着近年我国经济的高速腾飞,国民收入的迅猛增长,居民对保险的依赖程度逐步加强,我国保费收入也呈现逐年增长的趋势。因此,我们必要寻找科学的方法对保费收入进行准确的预测。由于保费收入具有明显的季节性特征,本文采用先定季节指数方法和X-12方法对保费收入进行分析,结果表明X-12方法更应用于此数据的季节性特征分析。为更好地预测保费收入的增长,本文首先建立季节性差分自回归滑动平均模型(SARIMA)和基于X-12乘法模型的自回归滑动平均模型(X-12-ARIMA乘法模型)以及基于X-12加法模型的自回归滑动平均模型(X-12-ARIMA加法模型)。通过此三个模型之间的比较,表明SARIMA模型和X-12-ARIMA乘法模型明显优于X-12-ARIMA加法模型。然后,本文提出SARIMA和X-12-ARIMA乘法模型相结合的组合模型对保费收入进行最终预测。在此组合模型中,我们采用粒子群优化算法对两模型的权重进行优化。最后我们对中国主要保险公司1999年1月至2013年6月间月度保费总收入时间序列进行实证分析,并对我国保费收入变化趋势进行预测,从而为我国保险业以及国家对于保险业的监管提供必要的支持。
[Abstract]:With the rapid development of Chinese economy and the rapid growth of national income in recent years, the reliance of residents on insurance is gradually strengthened, and the premium income of our country is also increasing year by year. Therefore, we need to find a scientific method to accurately predict premium income. Due to the obvious seasonal characteristics of premium income, the predefined seasonal index method and X-12 method are used to analyze the premium income. The results show that the X-12 method is more suitable for the seasonal characteristic analysis of this data. In order to better predict the growth of premium income, In this paper, we first establish seasonal differential autoregressive moving average model (SARIMA), autoregressive moving average model (X-12-ARIMA multiplication model) based on X-12 multiplication model and autoregressive sliding model based on X-12 addition model. Dynamic average model (X-12-ARIMA addition model). The comparison between the three models shows that the SARIMA model and the X-12-ARIMA multiplication model are obviously superior to the X-12-ARIMA addition model. Then, the combined model of SARIMA and X-12-ARIMA multiplication model is proposed to predict the premium income. In this combined model, particle swarm optimization algorithm is used to optimize the weights of the two models. Finally, we analyze the time series of monthly premium income of major Chinese insurance companies from January 1999 to June 2013, and predict the trend of premium income in China. So as to provide the necessary support for the insurance industry and the national supervision of the insurance industry.
【学位授予单位】：兰州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F224;F842
本文编号：2375267
[Abstract]:With the rapid development of Chinese economy and the rapid growth of national income in recent years, the reliance of residents on insurance is gradually strengthened, and the premium income of our country is also increasing year by year. Therefore, we need to find a scientific method to accurately predict premium income. Due to the obvious seasonal characteristics of premium income, the predefined seasonal index method and X-12 method are used to analyze the premium income. The results show that the X-12 method is more suitable for the seasonal characteristic analysis of this data. In order to better predict the growth of premium income, In this paper, we first establish seasonal differential autoregressive moving average model (SARIMA), autoregressive moving average model (X-12-ARIMA multiplication model) based on X-12 multiplication model and autoregressive sliding model based on X-12 addition model. Dynamic average model (X-12-ARIMA addition model). The comparison between the three models shows that the SARIMA model and the X-12-ARIMA multiplication model are obviously superior to the X-12-ARIMA addition model. Then, the combined model of SARIMA and X-12-ARIMA multiplication model is proposed to predict the premium income. In this combined model, particle swarm optimization algorithm is used to optimize the weights of the two models. Finally, we analyze the time series of monthly premium income of major Chinese insurance companies from January 1999 to June 2013, and predict the trend of premium income in China. So as to provide the necessary support for the insurance industry and the national supervision of the insurance industry.
【学位授予单位】：兰州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F224;F842
【参考文献】
相关期刊论文 前10条
1 包慧敏;王戎;;ARMA模型在人身险业增长预测中的应用研究[J];北方经济;2009年08期
2 尹成远;赵桂玲;周稳海;;中国人身保险保费收入的实证分析与预测研究[J];保险研究;2008年01期
3 张鸣芳;中国季度GDP季节调整分析[J];财经研究;2005年07期
4 傅庚，，唐小我，曾勇;广义递归方差倒数组合预测方法研究[J];电子科技大学学报;1995年02期
5 高春玲;;时间序列分解模型在寿险保费收入预测的应用[J];北方经贸;2010年06期
6 张云;高垒;;基于乘积季节模型的我国保费收入的预测研究[J];金融经济;2009年14期
7 张积林;;基于灰色理论的中国保险业保费规模预测[J];技术经济与管理研究;2010年01期
8 杨宁琳;;保险规模分析与预测方法述评[J];科技信息(科学教研);2007年33期
9 尹成远;李兆涛;;基于ARIMA模型对我国保费收入的预测[J];求索;2012年01期
10 王永宏;饶继广;;基于ARIMA模型自动预测我国保险行业保费收入的应用和实践[J];软件产业与工程;2010年06期
本文编号：2375267
本文链接：https://www.wllwen.com/jingjilunwen/bxjjlw/2375267.html
