计算机科学 ›› 2023, Vol. 50 ›› Issue (11): 41-48.doi: 10.11896/jsjkx.230500231
董红斌, 韩爽, 付强
DONG Hongbin, HAN Shuang, FU Qiang
摘要： 地理传感器时间序列具有复杂动态的语义时空相关性和地理时空相关性。尽管已经开发了各种深度学习模型用于时间序列预测,但很少有模型能专注于捕捉地理传感器时间序列内的多类型时空相关性。此外,同时预测多个传感器在未来某一时间步的值非常具有挑战性。为了解决上述问题,提出了一种自回归模型与深度神经网络的联合模型( Joint model of Autoregression and Deep Neural Network,J-ARDNN),用于处理地理传感器时间序列的多目标预测任务。在该模型中,空间模块用于捕捉不同序列间多类型空间的相关性,时间模块采用时间卷积网络来提取单个序列内的时间依赖关系。此外,还引入自回归模型来提高预测模型的鲁棒性。为了验证J-ARDNN模型的有效性和优越性,在不同领域的真实时间序列数据集上进行了充分的实验,结果表明,J-ARDNN模型的预测性能优于对比方法。
中图分类号:
| [1]LIANG Y,KE S,JUNZ,et al.GeoMAN:Multi-level Attention Networks for Geo-sensory Time Series Prediction[C]//Twenty-Seventh International Joint Conference on Artificial Intelligence.2018:3428-3434. [2]YIN C,DAI Q.A deep multivariate time series multi-step forecasting network[J].Applied Intelligence,2022,52(8):8956-8974. [3]CHEN D,CHEN L,ZHANG Y,et al.A Multiscale InteractiveRecurrent Network for Time- Series Forecasting[J].IEEE Transactions on Cybernetics,2022,52(9):8793-8803. [4]CANDANEDO L,FELDHEIM V.Data driven prediction mo-dels of energy use of appliances in a low-energy house[J].Energy and Buildings,2017,140(4):81-97. [5]MA Z,LIU S,GUO G,et al.Hybrid Attention Networks for Flow and Pressure Forecasting in Water Distribution Systems[J].IEEE Geoscience and Remote Sensing Letters,2022,19:1-5. [6]DU S D,LI T R,YANG Y,et al.A Sequence-to-Sequence Spatial-Temporal Attention Learning Model for Urban Traffic Flow Prediction[J].Journal of Computer Research and Development,2020,57(8):1715-1728. [7] LAI G,CHANG W,YANG Y,et al.Modeling long- and short-term temporal patterns with deep neural networks[C]//Proceedings of the 41st International ACM SIGIR Conference on Research and Development in Information Retrieval.New York:Association for Computing Machinery,2018:95-104. [8]SHIH S,SUN F,LEE H.Temporal pattern attention for multivariate time series forecasting[J].Machine Learning,2019,108:1421-1441. [9]CHANG Y,SUN F,WU Y,et al.A Memory-Network Based Solution for Multivariate Time-Series Forecasting[J].arXiv:1809.02105,2018. [10]HUANG S,WANG D,WU X,et al.DSANet:Dual Self-Attention Network for Multivariate Time Series Forecasting[C]//Proceedings ofthe 28th ACM International Conference on Information and Knowledge Management.New York:Association for Computing Machinery,2019:2129-2132. [11]WU Z,PAN S,LONG G,et al.Connecting the Dots:Multiva-riate Time Series Forecasting with Graph Neural Networks[C]//Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.New York:Association for Computing Machinery,2020:753-763. [12]JAIN A,KUMAR A M.Hybrid neural network models forhydrologic time series forecasting[J].Applied Soft Computing,2007,7(2):585-592. [13]HUANG G,WANG L.Hybrid Neural Network Models forHydrologic Time Series Forecasting Based on Genetic Algorithm[C]//Proceedings of the 4th International Joint Confe-rence on Computational Sciences and Optimization.New York:IEEE Press,2011:1347-1350. [14]FENG X,CHEN J,ZHANG Z,et al.State-of-charge estimation of lithium-ion battery based on clockwork recurrent neural network[J].Energy,2021,236:1-10. [15]ZHANG Y,PENG N,DAI M,et al.Memory-Gated Recurrent Networks[C]//Proceedings of the 35th AAAI Conference on Artificial Intelligence.Menlo Park:AAAI Press,2021:10956-10963. [16]MA Q,LIN Z,CHEN E,et al.Temporal pyramid recurrent neural network[C]//Proceedings of the 34th AAAI Conference on Artificial Intelligence.Menlo Park:AAAI Press,2020:5061-5068. [17]WANG X,ZHANG M,REN F.Sparse Gaussian ConditionalRandom Fields on top of Recurrent Neural Networks[C]//Proceedings of the 32nd AAAI Conference on Artificial Intelligence.Menlo Park:AAAI Press,2018:4219-4226. [18]MALDONADO S,GONZALEZ A,CRONE S.Automatic time series analysis for electric load forecasting via support vector regression[J].Applied Soft Computing,2019,83(3):1-10. [19]LIU J X,CHEN S.Non-Stationary Multi-variate Time SeriesPrediction with MIX Gated Unit[J].Journal of Computer Research and Development,2019,56(8):1642-1651. [20]LI Z X,LIU H Y.Combining Global and Sequential Patterns for Multivariate Time Series Forecasting[J].Chinese Journal of Computes,2023,46(1):70-84. [21]WANG J,SUN T,LIU B Y,et al.CLVSA:A Convolutional LSTM Based Variational Sequence-to-Sequence Model with Attention for Predicting Trends of Financial Markets[C]//Proceedings of International Joint Conference on Artificial Intelligence.San Francisco:Morgan Kaufmann.2021:3705-3711. [22]OZDEMIR A,BULUS K,ZOR K.Medium-to long-term nickel price forecasting using LSTM and GRU networks[J].Resources Policy,2022,78:1-10. [23]LIM B,ZOHREN S.Time Series Forecasting With Deep Lear-ning:A Survey[J].Philosophical Transactions of the Royal Society A,2021,379:1-12. [24]GUO T,LIN T,ANTULOV-FANTULIN N.Exploring Inter-pretable LSTM Neural Networks over Multi-Variable Data[C]//Proceedings of the 36th International Conference on Machine Learning.New York:ACM,2019:4424-4440. [25]VASWANI A,SHAZEER N,PARMAR N,et al.Attention is all you need[J].Advances in Neural Information Processing Systems,2017,30:5999-6009. [26]OORD A,DIELMAN S,ZEN H,et al.WaveNet:A Generative Model for Raw Audio[J].arXiv:1609.03499v2,2016. [27]GAN Z,LI C,ZHOU J,et al.Temporal convolutional networks interval prediction model for wind speed forecasting[J].Electric Power Systems Research,2021,191:1-10. [28]LI M,ZHU Z.Spatial-Temporal Fusion Graph Neural Networks for Traffic Flow Forecasting[C]//Proceedings of the 35th AAAI Conference on Artificial Intelligence.Menlo Park:AAAI Press,2021:4189-4196. [29]ZHANG G P.Time series forecasting using a hybrid ARIMA and neural network model[J].Neurocomputing,2003,50(6):159-175. [30]FRIFOLA R.Bayesian Time Series Learning with GaussianProcesses[D].Cambridgeshire:University of Cambridge,2015. |
|
||
首页