This article aims to provide an analysis for a time series data of gross domestic product (GDP) of the Sudan. An econometric time series model with macroeconomic variables is conducted. Since a non-stationary time ser...This article aims to provide an analysis for a time series data of gross domestic product (GDP) of the Sudan. An econometric time series model with macroeconomic variables is conducted. Since a non-stationary time series must be made stationary, some statistical tests are followed so that the time series become stationary series. After applying these tests, the time series became stationary and integrated of order I. Box-Jenkins procedure is used to determine ARMA. OLS is used to estimate the models parameters. Performances chosen ARIMA model are verified on the basis of classical statistical tests and forecasting. The model features are interpreted on the basis of standard measures of forecasting performance.展开更多
A generalized, structural, time series modeling framework was developed to analyze the monthly records of absolute surface temperature, one of the most important environmental parameters, using a deterministicstochast...A generalized, structural, time series modeling framework was developed to analyze the monthly records of absolute surface temperature, one of the most important environmental parameters, using a deterministicstochastic combined （DSC） approach. Although the development of the framework was based on the characterization of the variation patterns of a global dataset, the methodology could be applied to any monthly absolute temperature record. Deterministic processes were used to characterize the variation patterns of the global trend and the cyclic oscillations of the temperature signal, involving polynomial functions and the Fourier method, respectively, while stochastic processes were employed to account for any remaining patterns in the temperature signal, involving seasonal autoregressive integrated moving average （SARIMA） models. A prediction of the monthly global surface temperature during the second decade of the 21st century using the DSC model shows that the global temperature will likely continue to rise at twice the average rate of the past 150 years. The evaluation of prediction accuracy shows that DSC models perform systematically well against selected models of other authors, suggesting that DSC models, when coupled with other ecoenvironmental models, can be used as a supplemental tool for short-term （10-year） environmental planning and decision making.展开更多
探讨Mann-Kendall检验法、差分自回归移动平均模型(Autoregressive Mobile Average Model,ARIMA)与长短期记忆神经网络(Long and Short-Term Memory,LSTM)的组合模型在《中国的核安全》、《核安全年报》中核电厂运行事件数量的应用,对...探讨Mann-Kendall检验法、差分自回归移动平均模型(Autoregressive Mobile Average Model,ARIMA)与长短期记忆神经网络(Long and Short-Term Memory,LSTM)的组合模型在《中国的核安全》、《核安全年报》中核电厂运行事件数量的应用,对运行事件数量进行趋势分析、突变分析和预测并证明模型的适用性。利用1991⁓2018年核电厂运行事件数量,使用R软件建立ARIMA(2,1,2)模型,得到运行事件数量的线性部分;建立LSTM模型,对偏差序列进行预测,得到运行事件数量的非线性部分;最后建立ARIMA和LSTM组合模型,利用组合模型对运行事件数量进行预测,并根据实测数据对预测结果进行对比验证。实验结果表明:ARIMA和LSTM组合模型可较好地拟合运行事件数量时间序列,并修正单一模型的误差,有效提高预测精度3%,且得到的2019⁓2020年核电厂运行事件数量预测值与《核安全年报》相近。展开更多
文摘This article aims to provide an analysis for a time series data of gross domestic product (GDP) of the Sudan. An econometric time series model with macroeconomic variables is conducted. Since a non-stationary time series must be made stationary, some statistical tests are followed so that the time series become stationary series. After applying these tests, the time series became stationary and integrated of order I. Box-Jenkins procedure is used to determine ARMA. OLS is used to estimate the models parameters. Performances chosen ARIMA model are verified on the basis of classical statistical tests and forecasting. The model features are interpreted on the basis of standard measures of forecasting performance.
基金This research was supported by the Ministry of Science and Technology of China，National Basic Research Program of China (Grant No.2010CB951504).The authors acknowledge support from the Flemish Interuniversity Council，the Ghent University Laboratory of Soil Science for the writing of this paper
文摘A generalized, structural, time series modeling framework was developed to analyze the monthly records of absolute surface temperature, one of the most important environmental parameters, using a deterministicstochastic combined （DSC） approach. Although the development of the framework was based on the characterization of the variation patterns of a global dataset, the methodology could be applied to any monthly absolute temperature record. Deterministic processes were used to characterize the variation patterns of the global trend and the cyclic oscillations of the temperature signal, involving polynomial functions and the Fourier method, respectively, while stochastic processes were employed to account for any remaining patterns in the temperature signal, involving seasonal autoregressive integrated moving average （SARIMA） models. A prediction of the monthly global surface temperature during the second decade of the 21st century using the DSC model shows that the global temperature will likely continue to rise at twice the average rate of the past 150 years. The evaluation of prediction accuracy shows that DSC models perform systematically well against selected models of other authors, suggesting that DSC models, when coupled with other ecoenvironmental models, can be used as a supplemental tool for short-term （10-year） environmental planning and decision making.
文摘探讨Mann-Kendall检验法、差分自回归移动平均模型(Autoregressive Mobile Average Model,ARIMA)与长短期记忆神经网络(Long and Short-Term Memory,LSTM)的组合模型在《中国的核安全》、《核安全年报》中核电厂运行事件数量的应用,对运行事件数量进行趋势分析、突变分析和预测并证明模型的适用性。利用1991⁓2018年核电厂运行事件数量,使用R软件建立ARIMA(2,1,2)模型,得到运行事件数量的线性部分;建立LSTM模型,对偏差序列进行预测,得到运行事件数量的非线性部分;最后建立ARIMA和LSTM组合模型,利用组合模型对运行事件数量进行预测,并根据实测数据对预测结果进行对比验证。实验结果表明:ARIMA和LSTM组合模型可较好地拟合运行事件数量时间序列,并修正单一模型的误差,有效提高预测精度3%,且得到的2019⁓2020年核电厂运行事件数量预测值与《核安全年报》相近。
