Fuzzy sets theory cannot describe the neutrality degreeof data, which has largely limited the objectivity of fuzzy time seriesin uncertain data forecasting. With this regard, a multi-factor highorderintuitionistic fuz...Fuzzy sets theory cannot describe the neutrality degreeof data, which has largely limited the objectivity of fuzzy time seriesin uncertain data forecasting. With this regard, a multi-factor highorderintuitionistic fuzzy time series forecasting model is built. Inthe new model, a fuzzy clustering algorithm is used to get unequalintervals, and a more objective technique for ascertaining membershipand non-membership functions of the intuitionistic fuzzy setis proposed. On these bases, forecast rules based on multidimensionalintuitionistic fuzzy modus ponens inference are established.Finally, contrast experiments on the daily mean temperature ofBeijing are carried out, which show that the novel model has aclear advantage of improving the forecast accuracy.展开更多
Traditional studies on potential yield mainly referred to attainable yield： the maximum yield which could be reached by a crop in a given environment. The new concept of crop yield under average climate conditions wa...Traditional studies on potential yield mainly referred to attainable yield： the maximum yield which could be reached by a crop in a given environment. The new concept of crop yield under average climate conditions was defined in this paper, which was affected by advancement of science and technology. Based on the new concept of crop yield, the time series techniques relying on past yield data was employed to set up a forecasting model. The model was tested by using average grain yields of Liaoning Province in China from 1949 to 2005. The testing combined dynamic n-choosing and micro tendency rectification, and an average forecasting error was 1.24%. In the trend line of yield change, and then a yield turning point might occur, in which case the inflexion model was used to solve the problem of yield turn point.展开更多
The grey forecasting model has been successfully applied to many fields. However, the precision of GM(1,1) model is not high. In order to remove the seasonal fluctuations in monitoring series before building GM(1,1) m...The grey forecasting model has been successfully applied to many fields. However, the precision of GM(1,1) model is not high. In order to remove the seasonal fluctuations in monitoring series before building GM(1,1) model, the forecasting series of GM(1,1) was built, and an inverse process was used to resume the seasonal fluctuations. Two deseasonalization methods were presented , i.e., seasonal index-based deseasonalization and standard normal distribution-based deseasonalization. They were combined with the GM(1,1) model to form hybrid grey models. A simple but practical method to further improve the forecasting results was also suggested. For comparison, a conventional periodic function model was investigated. The concept and algorithms were tested with four years monthly monitoring data. The results show that on the whole the seasonal index-GM(1,1) model outperform the conventional periodic function model and the conventional periodic function model outperform the SND-GM(1,1) model. The mean Absolute error and mean square error of seasonal index-GM(1,1) are 30.69% and 54.53% smaller than that of conventional periodic function model, respectively. The high accuracy, straightforward and easy implementation natures of the proposed hybrid seasonal index-grey model make it a powerful analysis technique for seasonal monitoring series.展开更多
In order to effectively analyse the multivariate time series data of complex process,a generic reconstruction technology based on reduction theory of rough sets was proposed,Firstly,the phase space of multivariate tim...In order to effectively analyse the multivariate time series data of complex process,a generic reconstruction technology based on reduction theory of rough sets was proposed,Firstly,the phase space of multivariate time series was originally reconstructed by a classical reconstruction technology.Then,the original decision-table of rough set theory was set up according to the embedding dimensions and time-delays of the original reconstruction phase space,and the rough set reduction was used to delete the redundant dimensions and irrelevant variables and to reconstruct the generic phase space,Finally,the input vectors for the prediction of multivariate time series were extracted according to generic reconstruction results to identify the parameters of prediction model.Verification results show that the developed reconstruction method leads to better generalization ability for the prediction model and it is feasible and worthwhile for application.展开更多
Based on discussion on the theories of support vector machines （SVM）, an one-step prediction model for time series prediction is presented, wherein the chaos theory is incorporated. Chaotic character of the time ser...Based on discussion on the theories of support vector machines （SVM）, an one-step prediction model for time series prediction is presented, wherein the chaos theory is incorporated. Chaotic character of the time series is taken into account in the prediction procedure; parameters of reconstruction-detay and embedding-dimension for phase-space reconstruction are calculated in light of mutual-information and false-nearest-neighbor method, respectively. Precision and functionality have been demonstrated by the experimental results on the basis of the prediction of Lorenz chaotic time series.展开更多
传统时序预测模型通常仅关注捕捉复杂时序中的趋势和模式,而忽略了变量间的相互作用,限制了该模型在复杂时序预测中应用.提出一种Dualformer双模型并联方案,该模型并联iTransformer(inverted transformer)和PatchTST(patch time series ...传统时序预测模型通常仅关注捕捉复杂时序中的趋势和模式,而忽略了变量间的相互作用,限制了该模型在复杂时序预测中应用.提出一种Dualformer双模型并联方案,该模型并联iTransformer(inverted transformer)和PatchTST(patch time series transformer),通过激活函数替代前馈神经网络,并通过多层感知机计算输出结果.Dualformer利用注意力机制同时捕捉复杂时序中的时间维度和变量维度信息,关注时间趋势与多变量交互.实验结果显示,Dualformer在复杂时序预测效果上显著优于对比模型iTransformer、PatchTST和DLinear(decomposition linear),在实际应用中可显著提高复杂时序预测的准确度,具有广泛应用前景.展开更多
深度学习是解决时间序列分类(Time series classification,TSC)问题的主要途径之一.然而,基于深度学习的TSC模型易受到对抗样本攻击,从而导致模型分类准确率大幅度降低.为此,研究了TSC模型的对抗攻击防御问题,设计了集成对抗训练(Advers...深度学习是解决时间序列分类(Time series classification,TSC)问题的主要途径之一.然而,基于深度学习的TSC模型易受到对抗样本攻击,从而导致模型分类准确率大幅度降低.为此,研究了TSC模型的对抗攻击防御问题,设计了集成对抗训练(Adversarial training,AT)防御方法.首先,设计了一种针对TSC模型的集成对抗训练防御框架,通过多种TSC模型和攻击方式生成对抗样本,并用于训练目标模型.其次,在生成对抗样本的过程中,设计了基于Shapelets的局部扰动算法,并结合动量迭代的快速梯度符号法(Momentum iterative fast gradient sign method,MI-FGSM),实现了有效的白盒攻击.同时,使用知识蒸馏(Knowledge distillation,KD)和基于沃瑟斯坦距离的生成对抗网络(Wasserstein generative adversarial network,WGAN)设计了针对替代模型的黑盒对抗攻击方法,实现了攻击者对目标模型未知时的有效攻击.在此基础上,在对抗训练损失函数中添加Kullback-Leibler(KL)散度约束,进一步提升了模型鲁棒性.最后,在多变量时间序列分类数据集UEA上验证了所提方法的有效性.展开更多
基金supported by the National Natural Science Foundation of China(61309022)
文摘Fuzzy sets theory cannot describe the neutrality degreeof data, which has largely limited the objectivity of fuzzy time seriesin uncertain data forecasting. With this regard, a multi-factor highorderintuitionistic fuzzy time series forecasting model is built. Inthe new model, a fuzzy clustering algorithm is used to get unequalintervals, and a more objective technique for ascertaining membershipand non-membership functions of the intuitionistic fuzzy setis proposed. On these bases, forecast rules based on multidimensionalintuitionistic fuzzy modus ponens inference are established.Finally, contrast experiments on the daily mean temperature ofBeijing are carried out, which show that the novel model has aclear advantage of improving the forecast accuracy.
基金Supported by Agricultural Poor-helping Monopoly of Graduate University of Chinese Academy of Science (40641002)
文摘Traditional studies on potential yield mainly referred to attainable yield： the maximum yield which could be reached by a crop in a given environment. The new concept of crop yield under average climate conditions was defined in this paper, which was affected by advancement of science and technology. Based on the new concept of crop yield, the time series techniques relying on past yield data was employed to set up a forecasting model. The model was tested by using average grain yields of Liaoning Province in China from 1949 to 2005. The testing combined dynamic n-choosing and micro tendency rectification, and an average forecasting error was 1.24%. In the trend line of yield change, and then a yield turning point might occur, in which case the inflexion model was used to solve the problem of yield turn point.
文摘The grey forecasting model has been successfully applied to many fields. However, the precision of GM(1,1) model is not high. In order to remove the seasonal fluctuations in monitoring series before building GM(1,1) model, the forecasting series of GM(1,1) was built, and an inverse process was used to resume the seasonal fluctuations. Two deseasonalization methods were presented , i.e., seasonal index-based deseasonalization and standard normal distribution-based deseasonalization. They were combined with the GM(1,1) model to form hybrid grey models. A simple but practical method to further improve the forecasting results was also suggested. For comparison, a conventional periodic function model was investigated. The concept and algorithms were tested with four years monthly monitoring data. The results show that on the whole the seasonal index-GM(1,1) model outperform the conventional periodic function model and the conventional periodic function model outperform the SND-GM(1,1) model. The mean Absolute error and mean square error of seasonal index-GM(1,1) are 30.69% and 54.53% smaller than that of conventional periodic function model, respectively. The high accuracy, straightforward and easy implementation natures of the proposed hybrid seasonal index-grey model make it a powerful analysis technique for seasonal monitoring series.
基金Project(61025015) supported by the National Natural Science Funds for Distinguished Young Scholars of ChinaProject(21106036) supported by the National Natural Science Foundation of China+2 种基金Project(200805331103) supported by Research Fund for the Doctoral Program of Higher Education of ChinaProject(NCET-08-0576) supported by Program for New Century Excellent Talents in Universities of ChinaProject(11B038) supported by Scientific Research Fund for the Excellent Youth Scholars of Hunan Provincial Education Department,China
文摘In order to effectively analyse the multivariate time series data of complex process,a generic reconstruction technology based on reduction theory of rough sets was proposed,Firstly,the phase space of multivariate time series was originally reconstructed by a classical reconstruction technology.Then,the original decision-table of rough set theory was set up according to the embedding dimensions and time-delays of the original reconstruction phase space,and the rough set reduction was used to delete the redundant dimensions and irrelevant variables and to reconstruct the generic phase space,Finally,the input vectors for the prediction of multivariate time series were extracted according to generic reconstruction results to identify the parameters of prediction model.Verification results show that the developed reconstruction method leads to better generalization ability for the prediction model and it is feasible and worthwhile for application.
文摘Based on discussion on the theories of support vector machines （SVM）, an one-step prediction model for time series prediction is presented, wherein the chaos theory is incorporated. Chaotic character of the time series is taken into account in the prediction procedure; parameters of reconstruction-detay and embedding-dimension for phase-space reconstruction are calculated in light of mutual-information and false-nearest-neighbor method, respectively. Precision and functionality have been demonstrated by the experimental results on the basis of the prediction of Lorenz chaotic time series.
文摘传统时序预测模型通常仅关注捕捉复杂时序中的趋势和模式,而忽略了变量间的相互作用,限制了该模型在复杂时序预测中应用.提出一种Dualformer双模型并联方案,该模型并联iTransformer(inverted transformer)和PatchTST(patch time series transformer),通过激活函数替代前馈神经网络,并通过多层感知机计算输出结果.Dualformer利用注意力机制同时捕捉复杂时序中的时间维度和变量维度信息,关注时间趋势与多变量交互.实验结果显示,Dualformer在复杂时序预测效果上显著优于对比模型iTransformer、PatchTST和DLinear(decomposition linear),在实际应用中可显著提高复杂时序预测的准确度,具有广泛应用前景.
