论文部分内容阅读
目的:探索经验模态分解(EMD)在个人现金卫生支出长期趋势分析中的适用性,为我国个人现金卫生支出增长规律及影响因素研究提供新的思路。方法:提取1978—2013年中华人民共和国统计年鉴中个人现金卫生支出环比增长数据,运用EMD法将时间序列分解为不同本征模函数(IMF)和残差(RES),计算不同IMF的方差贡献率及波动周期。结果:以个人现金卫生支出环比增长速度为原始序列,成功分解出了4个IMF和RES。各IMF及RES的方差贡献率依次为44.58%,23.82%,6.12%,11.98%和13.49%,各IMF波动周期分别为3.0年、5.0年、11.7年和23.3年,分解结果不存在明显的端点效应和模态混叠现象。结论:EMD方法适用于个人现金卫生支出时间序列分解,分解结果可为进一步的预测及影响因素研究提供了基础。
Objective: To explore the applicability of Empirical Mode Decomposition (EMD) in the long-term trend analysis of personal cash health expenditure, and to provide new ideas for the research on the law of growth of personal cash health expenditure and its influencing factors. Methods: Extracting the growth of individual cash health expenditure in the Statistical Yearbook of the People’s Republic of China from 1978 to 2013, using the EMD method to decompose the time series into different IMFs and residuals (RES), and calculating the variance contribution of different IMFs Rate and fluctuation cycle. Results: With the growth rate of personal cash health expenditure as the original sequence, four IMFs and RESs were successfully decomposed. The variance contribution rates of IMFs and RESs were 44.58%, 23.82%, 6.12%, 11.98% and 13.49%, respectively. The volatility periods of IMFs were 3.0 years, 5.0 years, 11.7 years and 23.3 years respectively. There was no obvious endpoint Effect and modal aliasing phenomenon. Conclusion: The EMD method is suitable for the time series decomposition of individual cash health expenditure. The result of decomposition can provide the foundation for further forecasting and influencing factors.