Time series forecasting is essential for generating predictive insights across various domains, including healthcare, finance, and energy. This study focuses on forecasting patient health data by comparing the perform...Time series forecasting is essential for generating predictive insights across various domains, including healthcare, finance, and energy. This study focuses on forecasting patient health data by comparing the performance of traditional linear time series models, namely Autoregressive Integrated Moving Average (ARIMA), Seasonal ARIMA, and Moving Average (MA) against neural network architectures. The primary goal is to evaluate the effectiveness of these models in predicting healthcare outcomes using patient records, specifically the Cancerpatient.xlsx dataset, which tracks variables such as patient age, symptoms, genetic risk factors, and environmental exposures over time. The proposed strategy involves training each model on historical patient data to predict age progression and other related health indicators, with performance evaluated using Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) metrics. Our findings reveal that neural networks consistently outperform ARIMA and SARIMA by capturing non-linear patterns and complex temporal dependencies within the dataset, resulting in lower forecasting errors. This research highlights the potential of neural networks to enhance predictive accuracy in healthcare applications, supporting better resource allocation, patient monitoring, and long-term health outcome predictions.展开更多
Long-term time series forecasting stands as a crucial research domain within the realm of automated machine learning(AutoML).At present,forecasting,whether rooted in machine learning or statistical learning,typically ...Long-term time series forecasting stands as a crucial research domain within the realm of automated machine learning(AutoML).At present,forecasting,whether rooted in machine learning or statistical learning,typically relies on expert input and necessitates substantial manual involvement.This manual effort spans model development,feature engineering,hyper-parameter tuning,and the intricate construction of time series models.The complexity of these tasks renders complete automation unfeasible,as they inherently demand human intervention at multiple junctures.To surmount these challenges,this article proposes leveraging Long Short-Term Memory,which is the variant of Recurrent Neural Networks,harnessing memory cells and gating mechanisms to facilitate long-term time series prediction.However,forecasting accuracy by particular neural network and traditional models can degrade significantly,when addressing long-term time-series tasks.Therefore,our research demonstrates that this innovative approach outperforms the traditional Autoregressive Integrated Moving Average(ARIMA)method in forecasting long-term univariate time series.ARIMA is a high-quality and competitive model in time series prediction,and yet it requires significant preprocessing efforts.Using multiple accuracy metrics,we have evaluated both ARIMA and proposed method on the simulated time-series data and real data in both short and long term.Furthermore,our findings indicate its superiority over alternative network architectures,including Fully Connected Neural Networks,Convolutional Neural Networks,and Nonpooling Convolutional Neural Networks.Our AutoML approach enables non-professional to attain highly accurate and effective time series forecasting,and can be widely applied to various domains,particularly in business and finance.展开更多
Ocean temperature is an important physical variable in marine ecosystems,and ocean temperature prediction is an important research objective in ocean-related fields.Currently,one of the commonly used methods for ocean...Ocean temperature is an important physical variable in marine ecosystems,and ocean temperature prediction is an important research objective in ocean-related fields.Currently,one of the commonly used methods for ocean temperature prediction is based on data-driven,but research on this method is mostly limited to the sea surface,with few studies on the prediction of internal ocean temperature.Existing graph neural network-based methods usually use predefined graphs or learned static graphs,which cannot capture the dynamic associations among data.In this study,we propose a novel dynamic spatiotemporal graph neural network(DSTGN)to predict threedimensional ocean temperature(3D-OT),which combines static graph learning and dynamic graph learning to automatically mine two unknown dependencies between sequences based on the original 3D-OT data without prior knowledge.Temporal and spatial dependencies in the time series were then captured using temporal and graph convolutions.We also integrated dynamic graph learning,static graph learning,graph convolution,and temporal convolution into an end-to-end framework for 3D-OT prediction using time-series grid data.In this study,we conducted prediction experiments using high-resolution 3D-OT from the Copernicus global ocean physical reanalysis,with data covering the vertical variation of temperature from the sea surface to 1000 m below the sea surface.We compared five mainstream models that are commonly used for ocean temperature prediction,and the results showed that the method achieved the best prediction results at all prediction scales.展开更多
This paper presents a comparative study of ARIMA and Neural Network AutoRegressive (NNAR) models for time series forecasting. The study focuses on simulated data generated using ARIMA(1, 1, 0) and applies both models ...This paper presents a comparative study of ARIMA and Neural Network AutoRegressive (NNAR) models for time series forecasting. The study focuses on simulated data generated using ARIMA(1, 1, 0) and applies both models for training and forecasting. Model performance is evaluated using MSE, AIC, and BIC. The models are further applied to neonatal mortality data from Saudi Arabia to assess their predictive capabilities. The results indicate that the NNAR model outperforms ARIMA in both training and forecasting.展开更多
Time series analysis plays an important role in hydrologic forecasting,while the key to this analysis is to establish a proper model.This paper presents a time series neural network model with back propagation proced...Time series analysis plays an important role in hydrologic forecasting,while the key to this analysis is to establish a proper model.This paper presents a time series neural network model with back propagation procedure for hydrologic forecasting.Free from the disadvantages of previous models,the model can be parallel to operate information flexibly and rapidly.It excels in the ability of nonlinear mapping and can learn and adjust by itself,which gives the model a possibility to describe the complex nonlinear hydrologic process.By using directly a training process based on a set of previous data, the model can forecast the time series of stream flow.Moreover,two practical examples were used to test the performance of the time series neural network model.Results confirm that the model is efficient and feasible.展开更多
Time series forecasting plays an important role in various fields, such as energy, finance, transport, and weather. Temporal convolutional networks (TCNs) based on dilated causal convolution have been widely used in t...Time series forecasting plays an important role in various fields, such as energy, finance, transport, and weather. Temporal convolutional networks (TCNs) based on dilated causal convolution have been widely used in time series forecasting. However, two problems weaken the performance of TCNs. One is that in dilated casual convolution, causal convolution leads to the receptive fields of outputs being concentrated in the earlier part of the input sequence, whereas the recent input information will be severely lost. The other is that the distribution shift problem in time series has not been adequately solved. To address the first problem, we propose a subsequence-based dilated convolution method (SDC). By using multiple convolutional filters to convolve elements of neighboring subsequences, the method extracts temporal features from a growing receptive field via a growing subsequence rather than a single element. Ultimately, the receptive field of each output element can cover the whole input sequence. To address the second problem, we propose a difference and compensation method (DCM). The method reduces the discrepancies between and within the input sequences by difference operations and then compensates the outputs for the information lost due to difference operations. Based on SDC and DCM, we further construct a temporal subsequence-based convolutional network with difference (TSCND) for time series forecasting. The experimental results show that TSCND can reduce prediction mean squared error by 7.3% and save runtime, compared with state-of-the-art models and vanilla TCN.展开更多
In the Industrial Internet of Things(IIoT),sensors generate time series data to reflect the working state.When the systems are attacked,timely identification of outliers in time series is critical to ensure security.A...In the Industrial Internet of Things(IIoT),sensors generate time series data to reflect the working state.When the systems are attacked,timely identification of outliers in time series is critical to ensure security.Although many anomaly detection methods have been proposed,the temporal correlation of the time series over the same sensor and the state(spatial)correlation between different sensors are rarely considered simultaneously in these methods.Owing to the superior capability of Transformer in learning time series features.This paper proposes a time series anomaly detection method based on a spatial-temporal network and an improved Transformer.Additionally,the methods based on graph neural networks typically include a graph structure learning module and an anomaly detection module,which are interdependent.However,in the initial phase of training,since neither of the modules has reached an optimal state,their performance may influence each other.This scenario makes the end-to-end training approach hard to effectively direct the learning trajectory of each module.This interdependence between the modules,coupled with the initial instability,may cause the model to find it hard to find the optimal solution during the training process,resulting in unsatisfactory results.We introduce an adaptive graph structure learning method to obtain the optimal model parameters and graph structure.Experiments on two publicly available datasets demonstrate that the proposed method attains higher anomaly detection results than other methods.展开更多
The prediction for Multivariate Time Series(MTS)explores the interrelationships among variables at historical moments,extracts their relevant characteristics,and is widely used in finance,weather,complex industries an...The prediction for Multivariate Time Series(MTS)explores the interrelationships among variables at historical moments,extracts their relevant characteristics,and is widely used in finance,weather,complex industries and other fields.Furthermore,it is important to construct a digital twin system.However,existing methods do not take full advantage of the potential properties of variables,which results in poor predicted accuracy.In this paper,we propose the Adaptive Fused Spatial-Temporal Graph Convolutional Network(AFSTGCN).First,to address the problem of the unknown spatial-temporal structure,we construct the Adaptive Fused Spatial-Temporal Graph(AFSTG)layer.Specifically,we fuse the spatial-temporal graph based on the interrelationship of spatial graphs.Simultaneously,we construct the adaptive adjacency matrix of the spatial-temporal graph using node embedding methods.Subsequently,to overcome the insufficient extraction of disordered correlation features,we construct the Adaptive Fused Spatial-Temporal Graph Convolutional(AFSTGC)module.The module forces the reordering of disordered temporal,spatial and spatial-temporal dependencies into rule-like data.AFSTGCN dynamically and synchronously acquires potential temporal,spatial and spatial-temporal correlations,thereby fully extracting rich hierarchical feature information to enhance the predicted accuracy.Experiments on different types of MTS datasets demonstrate that the model achieves state-of-the-art single-step and multi-step performance compared with eight other deep learning models.展开更多
This paper proposes a co-evolutionary recurrent neural network （CERNN） for the multi-step-prediction of chaotic time series, it estimates the proper parameters of phase space reconstruction and optimizes the structu...This paper proposes a co-evolutionary recurrent neural network （CERNN） for the multi-step-prediction of chaotic time series, it estimates the proper parameters of phase space reconstruction and optimizes the structure of recurrent neural networks by coevolutionary strategy. The searching space was separated into two subspaces and the individuals are trained in a parallel computational procedure. It can dynamically combine the embedding method with the capability of recurrent neural network to incorporate past experience due to internal recurrence. The effectiveness of CERNN is evaluated by using three benchmark chaotic time series data sets： the Lorenz series, Mackey-Glass series and real-world sun spot series. The simulation results show that CERNN improves the performances of multi-step-prediction of chaotic time series.展开更多
In the real world, the inputs of many complicated systems are time-varying functions or processes. In order to predict the outputs of these systems with high speed and accuracy, this paper proposes a time series predi...In the real world, the inputs of many complicated systems are time-varying functions or processes. In order to predict the outputs of these systems with high speed and accuracy, this paper proposes a time series prediction model based on the wavelet process neural network, and develops the corresponding learning algorithm based on the expansion of the orthogonal basis functions. The effectiveness of the proposed time series prediction model and its learning algorithm is proved by the Macke-Glass time series prediction, and the comparative prediction results indicate that the proposed time series prediction model based on the wavelet process neural network seems to perform well and appears suitable for using as a good tool to predict the highly complex nonlinear time series.展开更多
To improve the diagnosis accuracy and self-adaptability of fatigue crack in ulterior place of the supporting shaft, time series and neural network are attempted to be applied in research on diag-nosing the fatigue cr...To improve the diagnosis accuracy and self-adaptability of fatigue crack in ulterior place of the supporting shaft, time series and neural network are attempted to be applied in research on diag-nosing the fatigue crack’s degree based on analyzing the vibration characteristics of the supporting shaft. By analyzing the characteristic parameter which is easy to be detected from the supporting shaft’s exterior, the time series model parameter which is hypersensitive to the situation of fatigue crack in ulterior place of the supporting shaft is the target input of neural network, and the fatigue crack’s degree value of supporting shaft is the output. The BP network model can be built and net-work can be trained after the structural parameters of network are selected. Furthermore, choosing the other two different group data can test the network. The test result will verify the validity of the BP network model. The result of experiment shows that the method of time series and neural network are effective to diagnose the occurrence and the development of the fatigue crack’s degree in ulterior place of the supporting shaft.展开更多
A nonlinear feedback term is introduced into the evaluation equation of weights of the backpropagation algorithm for neural network, the network becomes a chaotic one. For the purpose of that we can investigate how th...A nonlinear feedback term is introduced into the evaluation equation of weights of the backpropagation algorithm for neural network, the network becomes a chaotic one. For the purpose of that we can investigate how the different feedback terms affect the process of learning and forecasting, we use the model to forecast the nonlinear time series which is produced by Makey-Glass equation. By selecting the suitable feedback term, the system can escape from the local minima and converge to the global minimum or its approximate solutions, and the forecasting results are better than those of backpropagation algorithm.展开更多
On the assumption that random interruptions in the observation process are modeled by a sequence of independent Bernoulli random variables, we firstly generalize two kinds of nonlinear filtering methods with random in...On the assumption that random interruptions in the observation process are modeled by a sequence of independent Bernoulli random variables, we firstly generalize two kinds of nonlinear filtering methods with random interruption failures in the observation based on the extended Kalman filtering （EKF） and the unscented Kalman filtering （UKF）, which were shortened as GEKF and CUKF in this paper, respectively. Then the nonlinear filtering model is established by using the radial basis function neural network （RBFNN） prototypes and the network weights as state equation and the output of RBFNN to present the observation equation. Finally, we take the filtering problem under missing observed data as a special case of nonlinear filtering with random intermittent failures by setting each missing data to be zero without needing to pre-estimate the missing data, and use the GEKF-based RBFNN and the GUKF-based RBFNN to predict the ground radioactivity time series with missing data. Experimental results demonstrate that the prediction results of GUKF-based RBFNN accord well with the real ground radioactivity time series while the prediction results of GEKF-based RBFNN are divergent.展开更多
A new hybrid model which combines wavelets and Artificial Neural Network (ANN) called wavelet neural network (WNN) model was proposed in the current study and applied for time series modeling of river flow. The time s...A new hybrid model which combines wavelets and Artificial Neural Network (ANN) called wavelet neural network (WNN) model was proposed in the current study and applied for time series modeling of river flow. The time series of daily river flow of the Malaprabha River basin (Karnataka state, India) were analyzed by the WNN model. The observed time series are decomposed into sub-series using discrete wavelet transform and then appropriate sub-series is used as inputs to the neural network for forecasting hydrological variables. The hybrid model (WNN) was compared with the standard ANN and AR models. The WNN model was able to provide a good fit with the observed data, especially the peak values during the testing period. The benchmark results from WNN model applications showed that the hybrid model produced better results in estimating the hydrograph properties than the latter models (ANN and AR).展开更多
Artificial neural network (NN) is such a model as to imitate the structure and intelligence feature of human brain. It has strong nonlinear mapping function. To introduce NN into the study of earthquake prediction is ...Artificial neural network (NN) is such a model as to imitate the structure and intelligence feature of human brain. It has strong nonlinear mapping function. To introduce NN into the study of earthquake prediction is not only an extension of the application of artificial neural network model but also a new try for precursor observation to serve the earthquake prediction. In this paper, we analyzed the predictability of time series and gave a method of application of artificial neural network in forecasting earthquake precursor chaotic time series. Besides, taking the ground tilt observation of Jiangning and Xuzhou Station, the bulk strain observation of Liyang station as examples, we analyzed and forecasted their time series respectively. It is indicated that the precision of this method can meet the needs of practical task and therefore of great value in the application to the future practical earthquake analysis and prediction.展开更多
This paper discusses the modeling method of time series with neural network. In order to improve the adaptability of direct multi-step prediction models, this paper proposes a method of combining the temporal differen...This paper discusses the modeling method of time series with neural network. In order to improve the adaptability of direct multi-step prediction models, this paper proposes a method of combining the temporal differences methods with back-propagation algorithm for updating the parameters continuously on the basis of recent data. This method can make the neural network model fit the recent characteristic of the time series as close as possible, therefore improves the prediction accuracy. We built models and made predictions for the sunspot series. The prediction results of adaptive modeling method are better than that of non-adaptive modeling methods.展开更多
Molding and simulation of time series prediction based on dynamic neural network(NN) are studied. Prediction model for non-linear and time-varying system is proposed based on dynamic Jordan NN. Aiming at the intrinsic...Molding and simulation of time series prediction based on dynamic neural network(NN) are studied. Prediction model for non-linear and time-varying system is proposed based on dynamic Jordan NN. Aiming at the intrinsic defects of back-propagation (BP) algorithm that cannot update network weights incrementally, a hybrid algorithm combining the temporal difference (TD) method with BP algorithm to train Jordan NN is put forward. The proposed method is applied to predict the ash content of clean coal in jigging production real-time and multi-step. A practical example is also given and its application results indicate that the method has better performance than others and also offers a beneficial reference to the prediction of nonlinear time series.展开更多
Time series classification(TSC)has attracted a lot of attention for time series data mining tasks and has been applied in various fields.With the success of deep learning(DL)in computer vision recognition,people are s...Time series classification(TSC)has attracted a lot of attention for time series data mining tasks and has been applied in various fields.With the success of deep learning(DL)in computer vision recognition,people are starting to use deep learning to tackle TSC tasks.Quantum neural networks(QNN)have recently demonstrated their superiority over traditional machine learning in methods such as image processing and natural language processing,but research using quantum neural networks to handle TSC tasks has not received enough attention.Therefore,we proposed a learning framework based on multiple imaging and hybrid QNN(MIHQNN)for TSC tasks.We investigate the possibility of converting 1D time series to 2D images and classifying the converted images using hybrid QNN.We explored the differences between MIHQNN based on single time series imaging and MIHQNN based on the fusion of multiple time series imaging.Four quantum circuits were also selected and designed to study the impact of quantum circuits on TSC tasks.We tested our method on several standard datasets and achieved significant results compared to several current TSC methods,demonstrating the effectiveness of MIHQNN.This research highlights the potential of applying quantum computing to TSC and provides the theoretical and experimental background for future research.展开更多
Neural network and genetic algorithms are complementary technologies in the design of adaptive intelligent system. Neural network learns from scratch by adjusting the interconnections betweens layers. Genetic algorith...Neural network and genetic algorithms are complementary technologies in the design of adaptive intelligent system. Neural network learns from scratch by adjusting the interconnections betweens layers. Genetic algorithms are a popular computing framework that uses principals from natural population genetics to evolve solutions to problems. Various forecasting methods have been developed on the basis of neural network, but accuracy has been matter of concern in these forecasts. In neural network methods forecasted values depend to the choose of neural predictor structure, the number of the input, the lag. To remedy to these problem, in this paper, the authors are investing the applicability of an automatic design of a neural predictor realized by real Genetic Algorithms to predict the future value of a time series. The prediction method is tested by using meteorology time series that are daily and weekly mean temperatures in Melbourne, Australia, 1980-1990.展开更多
文摘Time series forecasting is essential for generating predictive insights across various domains, including healthcare, finance, and energy. This study focuses on forecasting patient health data by comparing the performance of traditional linear time series models, namely Autoregressive Integrated Moving Average (ARIMA), Seasonal ARIMA, and Moving Average (MA) against neural network architectures. The primary goal is to evaluate the effectiveness of these models in predicting healthcare outcomes using patient records, specifically the Cancerpatient.xlsx dataset, which tracks variables such as patient age, symptoms, genetic risk factors, and environmental exposures over time. The proposed strategy involves training each model on historical patient data to predict age progression and other related health indicators, with performance evaluated using Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) metrics. Our findings reveal that neural networks consistently outperform ARIMA and SARIMA by capturing non-linear patterns and complex temporal dependencies within the dataset, resulting in lower forecasting errors. This research highlights the potential of neural networks to enhance predictive accuracy in healthcare applications, supporting better resource allocation, patient monitoring, and long-term health outcome predictions.
文摘Long-term time series forecasting stands as a crucial research domain within the realm of automated machine learning(AutoML).At present,forecasting,whether rooted in machine learning or statistical learning,typically relies on expert input and necessitates substantial manual involvement.This manual effort spans model development,feature engineering,hyper-parameter tuning,and the intricate construction of time series models.The complexity of these tasks renders complete automation unfeasible,as they inherently demand human intervention at multiple junctures.To surmount these challenges,this article proposes leveraging Long Short-Term Memory,which is the variant of Recurrent Neural Networks,harnessing memory cells and gating mechanisms to facilitate long-term time series prediction.However,forecasting accuracy by particular neural network and traditional models can degrade significantly,when addressing long-term time-series tasks.Therefore,our research demonstrates that this innovative approach outperforms the traditional Autoregressive Integrated Moving Average(ARIMA)method in forecasting long-term univariate time series.ARIMA is a high-quality and competitive model in time series prediction,and yet it requires significant preprocessing efforts.Using multiple accuracy metrics,we have evaluated both ARIMA and proposed method on the simulated time-series data and real data in both short and long term.Furthermore,our findings indicate its superiority over alternative network architectures,including Fully Connected Neural Networks,Convolutional Neural Networks,and Nonpooling Convolutional Neural Networks.Our AutoML approach enables non-professional to attain highly accurate and effective time series forecasting,and can be widely applied to various domains,particularly in business and finance.
基金The National Key R&D Program of China under contract No.2021YFC3101603.
文摘Ocean temperature is an important physical variable in marine ecosystems,and ocean temperature prediction is an important research objective in ocean-related fields.Currently,one of the commonly used methods for ocean temperature prediction is based on data-driven,but research on this method is mostly limited to the sea surface,with few studies on the prediction of internal ocean temperature.Existing graph neural network-based methods usually use predefined graphs or learned static graphs,which cannot capture the dynamic associations among data.In this study,we propose a novel dynamic spatiotemporal graph neural network(DSTGN)to predict threedimensional ocean temperature(3D-OT),which combines static graph learning and dynamic graph learning to automatically mine two unknown dependencies between sequences based on the original 3D-OT data without prior knowledge.Temporal and spatial dependencies in the time series were then captured using temporal and graph convolutions.We also integrated dynamic graph learning,static graph learning,graph convolution,and temporal convolution into an end-to-end framework for 3D-OT prediction using time-series grid data.In this study,we conducted prediction experiments using high-resolution 3D-OT from the Copernicus global ocean physical reanalysis,with data covering the vertical variation of temperature from the sea surface to 1000 m below the sea surface.We compared five mainstream models that are commonly used for ocean temperature prediction,and the results showed that the method achieved the best prediction results at all prediction scales.
文摘This paper presents a comparative study of ARIMA and Neural Network AutoRegressive (NNAR) models for time series forecasting. The study focuses on simulated data generated using ARIMA(1, 1, 0) and applies both models for training and forecasting. Model performance is evaluated using MSE, AIC, and BIC. The models are further applied to neonatal mortality data from Saudi Arabia to assess their predictive capabilities. The results indicate that the NNAR model outperforms ARIMA in both training and forecasting.
文摘Time series analysis plays an important role in hydrologic forecasting,while the key to this analysis is to establish a proper model.This paper presents a time series neural network model with back propagation procedure for hydrologic forecasting.Free from the disadvantages of previous models,the model can be parallel to operate information flexibly and rapidly.It excels in the ability of nonlinear mapping and can learn and adjust by itself,which gives the model a possibility to describe the complex nonlinear hydrologic process.By using directly a training process based on a set of previous data, the model can forecast the time series of stream flow.Moreover,two practical examples were used to test the performance of the time series neural network model.Results confirm that the model is efficient and feasible.
基金supported by the National Key Research and Development Program of China(No.2018YFB2101300)the National Natural Science Foundation of China(Grant No.61871186)the Dean’s Fund of Engineering Research Center of Software/Hardware Co-Design Technology and Application,Ministry of Education(East China Normal University).
文摘Time series forecasting plays an important role in various fields, such as energy, finance, transport, and weather. Temporal convolutional networks (TCNs) based on dilated causal convolution have been widely used in time series forecasting. However, two problems weaken the performance of TCNs. One is that in dilated casual convolution, causal convolution leads to the receptive fields of outputs being concentrated in the earlier part of the input sequence, whereas the recent input information will be severely lost. The other is that the distribution shift problem in time series has not been adequately solved. To address the first problem, we propose a subsequence-based dilated convolution method (SDC). By using multiple convolutional filters to convolve elements of neighboring subsequences, the method extracts temporal features from a growing receptive field via a growing subsequence rather than a single element. Ultimately, the receptive field of each output element can cover the whole input sequence. To address the second problem, we propose a difference and compensation method (DCM). The method reduces the discrepancies between and within the input sequences by difference operations and then compensates the outputs for the information lost due to difference operations. Based on SDC and DCM, we further construct a temporal subsequence-based convolutional network with difference (TSCND) for time series forecasting. The experimental results show that TSCND can reduce prediction mean squared error by 7.3% and save runtime, compared with state-of-the-art models and vanilla TCN.
基金This work is partly supported by the National Key Research and Development Program of China(Grant No.2020YFB1805403)the National Natural Science Foundation of China(Grant No.62032002)the 111 Project(Grant No.B21049).
文摘In the Industrial Internet of Things(IIoT),sensors generate time series data to reflect the working state.When the systems are attacked,timely identification of outliers in time series is critical to ensure security.Although many anomaly detection methods have been proposed,the temporal correlation of the time series over the same sensor and the state(spatial)correlation between different sensors are rarely considered simultaneously in these methods.Owing to the superior capability of Transformer in learning time series features.This paper proposes a time series anomaly detection method based on a spatial-temporal network and an improved Transformer.Additionally,the methods based on graph neural networks typically include a graph structure learning module and an anomaly detection module,which are interdependent.However,in the initial phase of training,since neither of the modules has reached an optimal state,their performance may influence each other.This scenario makes the end-to-end training approach hard to effectively direct the learning trajectory of each module.This interdependence between the modules,coupled with the initial instability,may cause the model to find it hard to find the optimal solution during the training process,resulting in unsatisfactory results.We introduce an adaptive graph structure learning method to obtain the optimal model parameters and graph structure.Experiments on two publicly available datasets demonstrate that the proposed method attains higher anomaly detection results than other methods.
基金supported by the China Scholarship Council and the CERNET Innovation Project under grant No.20170111.
文摘The prediction for Multivariate Time Series(MTS)explores the interrelationships among variables at historical moments,extracts their relevant characteristics,and is widely used in finance,weather,complex industries and other fields.Furthermore,it is important to construct a digital twin system.However,existing methods do not take full advantage of the potential properties of variables,which results in poor predicted accuracy.In this paper,we propose the Adaptive Fused Spatial-Temporal Graph Convolutional Network(AFSTGCN).First,to address the problem of the unknown spatial-temporal structure,we construct the Adaptive Fused Spatial-Temporal Graph(AFSTG)layer.Specifically,we fuse the spatial-temporal graph based on the interrelationship of spatial graphs.Simultaneously,we construct the adaptive adjacency matrix of the spatial-temporal graph using node embedding methods.Subsequently,to overcome the insufficient extraction of disordered correlation features,we construct the Adaptive Fused Spatial-Temporal Graph Convolutional(AFSTGC)module.The module forces the reordering of disordered temporal,spatial and spatial-temporal dependencies into rule-like data.AFSTGCN dynamically and synchronously acquires potential temporal,spatial and spatial-temporal correlations,thereby fully extracting rich hierarchical feature information to enhance the predicted accuracy.Experiments on different types of MTS datasets demonstrate that the model achieves state-of-the-art single-step and multi-step performance compared with eight other deep learning models.
基金Project supported by the State Key Program of National Natural Science of China (Grant No 30230350)the Natural Science Foundation of Guangdong Province,China (Grant No 07006474)
文摘This paper proposes a co-evolutionary recurrent neural network （CERNN） for the multi-step-prediction of chaotic time series, it estimates the proper parameters of phase space reconstruction and optimizes the structure of recurrent neural networks by coevolutionary strategy. The searching space was separated into two subspaces and the individuals are trained in a parallel computational procedure. It can dynamically combine the embedding method with the capability of recurrent neural network to incorporate past experience due to internal recurrence. The effectiveness of CERNN is evaluated by using three benchmark chaotic time series data sets： the Lorenz series, Mackey-Glass series and real-world sun spot series. The simulation results show that CERNN improves the performances of multi-step-prediction of chaotic time series.
基金Project supported by the National Natural Science Foundation of China (Grant No 60572174)the Doctoral Fund of Ministry of Education of China (Grant No 20070213072)+2 种基金the 111 Project (Grant No B07018)the China Postdoctoral Science Foundation (Grant No 20070410264)the Development Program for Outstanding Young Teachers in Harbin Institute of Technology (Grant No HITQNJS.2007.010)
文摘In the real world, the inputs of many complicated systems are time-varying functions or processes. In order to predict the outputs of these systems with high speed and accuracy, this paper proposes a time series prediction model based on the wavelet process neural network, and develops the corresponding learning algorithm based on the expansion of the orthogonal basis functions. The effectiveness of the proposed time series prediction model and its learning algorithm is proved by the Macke-Glass time series prediction, and the comparative prediction results indicate that the proposed time series prediction model based on the wavelet process neural network seems to perform well and appears suitable for using as a good tool to predict the highly complex nonlinear time series.
基金This project is supported by National Natural Science Fundation of China (No. 50675066)Provincial Key Technologies R&D of Hunan, China (No. 05FJ2001)China Postdoctoral Science Foundation (No. 2005038006).
文摘To improve the diagnosis accuracy and self-adaptability of fatigue crack in ulterior place of the supporting shaft, time series and neural network are attempted to be applied in research on diag-nosing the fatigue crack’s degree based on analyzing the vibration characteristics of the supporting shaft. By analyzing the characteristic parameter which is easy to be detected from the supporting shaft’s exterior, the time series model parameter which is hypersensitive to the situation of fatigue crack in ulterior place of the supporting shaft is the target input of neural network, and the fatigue crack’s degree value of supporting shaft is the output. The BP network model can be built and net-work can be trained after the structural parameters of network are selected. Furthermore, choosing the other two different group data can test the network. The test result will verify the validity of the BP network model. The result of experiment shows that the method of time series and neural network are effective to diagnose the occurrence and the development of the fatigue crack’s degree in ulterior place of the supporting shaft.
文摘A nonlinear feedback term is introduced into the evaluation equation of weights of the backpropagation algorithm for neural network, the network becomes a chaotic one. For the purpose of that we can investigate how the different feedback terms affect the process of learning and forecasting, we use the model to forecast the nonlinear time series which is produced by Makey-Glass equation. By selecting the suitable feedback term, the system can escape from the local minima and converge to the global minimum or its approximate solutions, and the forecasting results are better than those of backpropagation algorithm.
基金Project supported by the State Key Program of the National Natural Science of China (Grant No. 60835004)the Natural Science Foundation of Jiangsu Province of China (Grant No. BK2009727)+1 种基金the Natural Science Foundation of Higher Education Institutions of Jiangsu Province of China (Grant No. 10KJB510004)the National Natural Science Foundation of China (Grant No. 61075028)
文摘On the assumption that random interruptions in the observation process are modeled by a sequence of independent Bernoulli random variables, we firstly generalize two kinds of nonlinear filtering methods with random interruption failures in the observation based on the extended Kalman filtering （EKF） and the unscented Kalman filtering （UKF）, which were shortened as GEKF and CUKF in this paper, respectively. Then the nonlinear filtering model is established by using the radial basis function neural network （RBFNN） prototypes and the network weights as state equation and the output of RBFNN to present the observation equation. Finally, we take the filtering problem under missing observed data as a special case of nonlinear filtering with random intermittent failures by setting each missing data to be zero without needing to pre-estimate the missing data, and use the GEKF-based RBFNN and the GUKF-based RBFNN to predict the ground radioactivity time series with missing data. Experimental results demonstrate that the prediction results of GUKF-based RBFNN accord well with the real ground radioactivity time series while the prediction results of GEKF-based RBFNN are divergent.
文摘A new hybrid model which combines wavelets and Artificial Neural Network (ANN) called wavelet neural network (WNN) model was proposed in the current study and applied for time series modeling of river flow. The time series of daily river flow of the Malaprabha River basin (Karnataka state, India) were analyzed by the WNN model. The observed time series are decomposed into sub-series using discrete wavelet transform and then appropriate sub-series is used as inputs to the neural network for forecasting hydrological variables. The hybrid model (WNN) was compared with the standard ANN and AR models. The WNN model was able to provide a good fit with the observed data, especially the peak values during the testing period. The benchmark results from WNN model applications showed that the hybrid model produced better results in estimating the hydrograph properties than the latter models (ANN and AR).
文摘Artificial neural network (NN) is such a model as to imitate the structure and intelligence feature of human brain. It has strong nonlinear mapping function. To introduce NN into the study of earthquake prediction is not only an extension of the application of artificial neural network model but also a new try for precursor observation to serve the earthquake prediction. In this paper, we analyzed the predictability of time series and gave a method of application of artificial neural network in forecasting earthquake precursor chaotic time series. Besides, taking the ground tilt observation of Jiangning and Xuzhou Station, the bulk strain observation of Liyang station as examples, we analyzed and forecasted their time series respectively. It is indicated that the precision of this method can meet the needs of practical task and therefore of great value in the application to the future practical earthquake analysis and prediction.
文摘This paper discusses the modeling method of time series with neural network. In order to improve the adaptability of direct multi-step prediction models, this paper proposes a method of combining the temporal differences methods with back-propagation algorithm for updating the parameters continuously on the basis of recent data. This method can make the neural network model fit the recent characteristic of the time series as close as possible, therefore improves the prediction accuracy. We built models and made predictions for the sunspot series. The prediction results of adaptive modeling method are better than that of non-adaptive modeling methods.
文摘Molding and simulation of time series prediction based on dynamic neural network(NN) are studied. Prediction model for non-linear and time-varying system is proposed based on dynamic Jordan NN. Aiming at the intrinsic defects of back-propagation (BP) algorithm that cannot update network weights incrementally, a hybrid algorithm combining the temporal difference (TD) method with BP algorithm to train Jordan NN is put forward. The proposed method is applied to predict the ash content of clean coal in jigging production real-time and multi-step. A practical example is also given and its application results indicate that the method has better performance than others and also offers a beneficial reference to the prediction of nonlinear time series.
基金Project supported by the National Natural Science Foundation of China (Grant Nos.61772295 and 61572270)the PHD foundation of Chongqing Normal University (Grant No.19XLB003)Chongqing Technology Foresight and Institutional Innovation Project (Grant No.cstc2021jsyjyzysbAX0011)。
文摘Time series classification(TSC)has attracted a lot of attention for time series data mining tasks and has been applied in various fields.With the success of deep learning(DL)in computer vision recognition,people are starting to use deep learning to tackle TSC tasks.Quantum neural networks(QNN)have recently demonstrated their superiority over traditional machine learning in methods such as image processing and natural language processing,but research using quantum neural networks to handle TSC tasks has not received enough attention.Therefore,we proposed a learning framework based on multiple imaging and hybrid QNN(MIHQNN)for TSC tasks.We investigate the possibility of converting 1D time series to 2D images and classifying the converted images using hybrid QNN.We explored the differences between MIHQNN based on single time series imaging and MIHQNN based on the fusion of multiple time series imaging.Four quantum circuits were also selected and designed to study the impact of quantum circuits on TSC tasks.We tested our method on several standard datasets and achieved significant results compared to several current TSC methods,demonstrating the effectiveness of MIHQNN.This research highlights the potential of applying quantum computing to TSC and provides the theoretical and experimental background for future research.
文摘Neural network and genetic algorithms are complementary technologies in the design of adaptive intelligent system. Neural network learns from scratch by adjusting the interconnections betweens layers. Genetic algorithms are a popular computing framework that uses principals from natural population genetics to evolve solutions to problems. Various forecasting methods have been developed on the basis of neural network, but accuracy has been matter of concern in these forecasts. In neural network methods forecasted values depend to the choose of neural predictor structure, the number of the input, the lag. To remedy to these problem, in this paper, the authors are investing the applicability of an automatic design of a neural predictor realized by real Genetic Algorithms to predict the future value of a time series. The prediction method is tested by using meteorology time series that are daily and weekly mean temperatures in Melbourne, Australia, 1980-1990.
