Multi-View Time Series Prediction Based on Consistent Functional Neural Process

doi:10.19678/j.issn.1000-3428.0069594

Abstract

Abstract:

In the field of multi-view time series prediction, the effective fusion of information from different views is an important and challenging issue. Existing multi-view time series prediction methods have limitations in capturing historical data trends and are often affected by the inconsistent distribution of multi-view information. To address these two problems, this study leverages the Functional Neural Process (FNP) framework and proposes a Consistent FNP (CFNP) framework. The CFNP framework is designed with two core modules: view random correlation graph module and view distribution alignment module. The view random correlation graph module assists in understanding and predicting current data by analyzing the distribution of historical data. The view distribution alignment module is dedicated to reducing the difference in probability distributions between different views and improving the time response of the model by imposing constraints in the potential space. Consequently, the model can capture the intrinsic correlation of sequences. On two public datasets, the CFNP framework improves the Root Mean Square Error (RMSE) by 14% and 5% compared to existing methods, proving that it can predict multi-view time series more accurately.

Key words: multi-view learning, time series prediction, probabilistic prediction, Functional Neural Process (FNP), consistency regularization

摘要：

在多视图时间序列预测领域, 如何有效融合来自不同视图的信息, 是一个重要且具有挑战性的问题。现有的多视图时序预测方法在捕获历史数据趋势方面存在局限性, 同时也常受到多视图信息分布不一致的影响。针对这两个问题, 基于功能神经过程(FNP)框架, 提出一种一致性功能神经过程(CFNP)框架。CFNP框架中包含两个核心模块: 视图随机相关图模块和视图分布对齐模块。视图随机相关图模块通过分析历史数据的分布, 辅助对当前数据的理解和预测; 而视图分布对齐模块致力于缩小不同视图间的概率分布差异, 通过在潜在空间中施加约束, 提高模型对时间序列内在关联性的捕捉能力。在两个公开数据集上的实验结果表明, 相比于现有方法, CFNP框架在均方根误差(RMSE)上性能提升分别为14%和5%, 证明此框架能够更准确地预测多视图时间序列。

关键词: 多视图学习, 时序预测, 概率预测, 功能神经过程, 一致性正则化

YANG Chunxia, JIANG Yao, ZHAI Xuetong, ZHOU Yuanyuan. Multi-View Time Series Prediction Based on Consistent Functional Neural Process[J]. Computer Engineering, 2025, 51(10): 111-120.

杨春霞, 蒋耀, 翟雪彤, 周媛媛. 基于一致性功能神经过程的多视图时序预测[J]. 计算机工程, 2025, 51(10): 111-120.

/ Recommend / Download Citations

URL: https://www.ecice06.com/EN/10.19678/j.issn.1000-3428.0069594

https://www.ecice06.com/EN/Y2025/V51/I10/111

Figures/Tables 9

Fig.1 The main flow of prediction model

Fig.2 View random correlation graph module

Fig.3 View distribution alignment module

Fig.4 Model optimal learning rate parameter experiment result

Fig.5 Effect of model hidden layer dimension and linear layer number on RMSE in ILI-Symptom dataset

References 33

1	LI L Y, YAN J C, YANG X K, et al. Learning interpretable deep state space model for probabilistic time series forecasting[C]//Proceedings of the 28th International Joint Conference on Artificial Intelligence. Macao, China: International Joint Conferences on Artificial Intelligence Organization, 2019: 2901-2908.
2	SALINAS D , FLUNKERT V , GASTHAUS J , et al. DeepAR: probabilistic forecasting with autoregressive recurrent networks. International Journal of Forecasting, 2020, 36 (3): 1181- 1191. doi: 10.1016/j.ijforecast.2019.07.001
3	HU W J, YANG Y, WANG J B, et al. Understanding electricity-theft behavior via multi-source data[C]//Proceedings of the Web Conference 2020. New York, USA: ACM Press, 2020: 2264-2274.
4	TRIRAT P, LEE J G. DF-TAR: a deep fusion network for citywide traffic accident risk prediction with dangerous driving behavior[C]//Proceedings of the Web Conference 2021. New York, USA: ACM Press, 2021: 1146-1156.
5	RAMCHANDANI A , FAN C , MOSTAFAVI A . DeepCOVIDNet: an interpretable deep learning model for predictive surveillance of COVID-19 using heterogeneous features and their interactions. IEEE Access, 2020, 8, 159915- 159930. doi: 10.1109/ACCESS.2020.3019989
6	CHEN C C, LIANG J J, MA F L, et al. UNITE: uncertainty-based health risk prediction leveraging multi-sourced data[C]//Proceedings of the Web Conference 2021. New York, USA: ACM Press, 2021: 217-226.
7	YAN X Q , HU S Z , MAO Y Q , et al. Deep multi-view learning methods: a review. Neurocomputing, 2021, 448, 106- 129. doi: 10.1016/j.neucom.2021.03.090
8	KAMARTHI H, KONG L K, RODRIGUEZ A, et al. CAMul: calibrated and accurate multi-view time-series forecasting[C]//Proceedings of the Web Conference 2022. New York, USA: ACM Press, 2022: 3174-3185.
9	QIN Y, SONG D J, CHEN H F, et al. A dual-stage attention-based recurrent neural network for time series prediction[EB/OL]. [2024-02-01]. https://arxiv.org/abs/1704.02971v4.
10	ADHIKARI B, XU X F, RAMAKRISHNAN N, et al. EpiDeep: exploiting embeddings for epidemic forecasting[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, USA: ACM Press, 2019: 577-586.
11	TANG X F , YAO H X , SUN Y W , et al. Joint modeling of local and global temporal dynamics for multivariate time series forecasting with missing values. Proceedings of the AAAI Conference on Artificial Intelligence, 2020, 34 (4): 5956- 5963. doi: 10.1609/aaai.v34i04.6056
12	TAN Q X , YE M , YANG B Y , et al. DATA-GRU: dual-attention time-aware gated recurrent unit for irregular multivariate time series. Proceedings of the AAAI Conference on Artificial Intelligence, 2020, 34 (1): 930- 937. doi: 10.1609/aaai.v34i01.5440
13	杨静, 陆铭华, 马洁琼, 等. 基于交替循环神经网络的水下防御态势预测方法. 计算机工程, 2023, 49 (9): 69- 78. doi: 10.19678/j.issn.1000-3428.0065183
	YANG J , LU M H , MA J Q , et al. Underwater defense posture prediction method based on alternating recurrent neural network. Computer Engineering, 2023, 49 (9): 69- 78. doi: 10.19678/j.issn.1000-3428.0065183
14	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. New York, USA: ACM Press, 2017: 6000-6010.
15	韩璐, 霍纬纲, 张永会, 等. 基于多尺度特征融合与双注意力机制的多元时间序列预测. 计算机工程, 2023, 49 (9): 99- 108. doi: 10.19678/j.issn.1000-3428.0065846
	HAN L , HUO W Q , ZHANG Y H , et al. Multivariate time series forecasting based on multi-scale feature fusion and dual-attention mechanism. Computer Engineering, 2023, 49 (9): 99- 108. doi: 10.19678/j.issn.1000-3428.0065846
16	刘杭, 殷歆, 陈杰, 等. 基于混合网络模型的多维时间序列预测. 计算机工程, 2023, 49 (1): 121- 129. doi: 10.19678/j.issn.1000-3428.0063718
	LIU H , YIN X , CHEN J , et al. Multi-dimensional time-series prediction based on hybrid network models. Computer Engineering, 2023, 49 (1): 121- 129. doi: 10.19678/j.issn.1000-3428.0063718
17	SALINAS D , FLUNKERT V , GASTHAUS J , et al. DeepAR: probabilistic forecasting with autoregressive recurrent networks. International Journal of Forecasting, 2020, 36 (3): 1181- 1191. doi: 10.1016/j.ijforecast.2019.07.001
18	KRISHNAN R , SHALIT U , SONTAG D . Structured inference networks for nonlinear state space models. Proceedings of the AAAI Conference on Artificial Intelligence, 2017, 31 (1): 2101- 2109.
19	LI L Y , YAN J C , ZHANG Y H , et al. Learning generative RNN-ODE for collaborative time-series and event sequence forecasting. IEEE Transactions on Knowledge and Data Engineering, 2023, 35 (7): 7118- 7137.
20	FRACCARO M, SØNDERBY S K, PAQUET U, et al. Sequential neural models with stochastic layers[C]//Proceedings of the 30th International Conference on Neural Information Processing Systems. Berlin, Germany: Springer, 2016: 2207-2215.
21	ZHENG X, ZAHEER M, AHMED A, et al. State space LSTM models with particle MCMC inference[EB/OL]. [2024-02-01]. https://arxiv.org/abs/1711.11179v1.
22	GARNELO M, SCHWARZ J, ROSENBAUM D, et al. Neural processes[EB/OL]. [2024-02-01]. https://arxiv.org/abs/1807.01622v1.
23	QIN S H, ZHU J C, QIN J, et al. Recurrent attentive neural process for sequential data[EB/OL]. [2024-02-01]. https://arxiv.org/abs/1910.09323v1.
24	LOUIZOS C, SHI X, SCHUTTE K, et al. The functional neural process[C]//Proceedings of the 33rd International Conference on Neural Information Processing Systems. Berlin, Germany: Springer, 2019: 8746-8757.
25	KAMARTHI H , KONG L , RODRIGUEZ A , et al. When in doubt: neural non-parametric uncertainty quantification for epidemic forecasting. Advances in Neural Information Processing Systems, 2021, 34, 19796- 19807.
26	ZHANG N , SUN S . Multiview unsupervised shapelet learning for multivariate time series clustering. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45 (4): 4981- 4996. doi: 10.1109/TPAMI.2022.3198411
27	LI S, LI Y L, FU Y. Multi-view time series classification: a discriminative bilinear projection approach[C]//Proceedings of the 25th ACM International on Conference on Information and Knowledge Management. New York, USA: ACM Press, 2016: 989-998.
28	EKAMBARAM V, MANGLIK K, MUKHERJEE S, et al. Attention based multi-modal new product sales time-series forecasting[C]//Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, USA: ACM Press, 2020: 3110-3118.
29	PANDEY A , MEULEMEESTER H , MOOR B , et al. Multi-view kernel PCA for time series forecasting. Neurocomputing, 2023, 554, 126639. doi: 10.1016/j.neucom.2023.126639
30	KUMAR R, YADAV S, DANIULAITYTE R, et al. eDarkFind: unsupervised multi-view learning for sybil account detection[C]//Proceedings of the Web Conference 2020. New York, USA: ACM Press, 2020: 1955-1965.
31	YAO H, WU F, KE J, et al. Deep multi-view spatial-temporal network for taxi demand prediction[C]//Proceedings of the 32nd AAAI Conference on Artificial Intelligence. [S. l. ]: AAAI press, 2018: 2588-2595.
32	ZHONG B, WANG P, PAN J, et al. TS-MVP: time-series representation learning by multi-view prototypical contrastive learning[C]//Proceedings of International Conference on Advanced Data Mining and Applications. Berlin, Germany: Springer, 2023: 278-292.
33	林彬. 基于多视图特征增强与融合的术中低血压预测研究[D]. 武汉: 华中科技大学, 2020.
	LIN B. Research on intraoperative hypotension prediction based on multi-view feature enhancement and fusion[D]. Wuhan: Huazhong University of Science and Technology, 2020. (in Chinese)

[1]	ZHANG Xiao, LI Deshi. Prediction Mechanism Based on Multi-dimensional Hidden States for Non-stationary Time Series [J]. Computer Engineering, 2025, 51(7): 68-77.
[2]	LI Ping, ZHANG Xueying, WANG Suzhe, LI Fenglian, ZHANG Hua. Medical Image Segmentation Based on Semi-Supervised Multi-Scale Consistency Learning [J]. Computer Engineering, 2025, 51(10): 295-307.
[3]	HU Aoran, CHEN Xiaohong. One-step Multi-view Clustering Based on Diversity and Consistency [J]. Computer Engineering, 2024, 50(5): 51-61.
[4]	LEI Jie, RAO Wenbi, YANG Yanchao, XIONG Shengwu. Pseudo-Label Object Detection Algorithm Based on Classification Uncertainty [J]. Computer Engineering, 2023, 49(1): 49-56.
[5]	LI Xiao, LU Xianling. ethod for Forecasting Short-Term Power Load Based on Dual-Stage Attention Mechanism and Gated Recurrent Unit Network [J]. Computer Engineering, 2022, 48(2): 291-296,305.
[6]	LIU Yanwen, ZHANG Jinxin, ZHANG Hongjie, JING Ling. Incomplete Multi-view Embedded Learning Method Based on Double Locality Preserving [J]. Computer Engineering, 2021, 47(6): 115-122,141.
[7]	ZHANG Enhao, CHEN Xiaohong, LIU Hong, ZHU Yulian. Overview of Multi-view Dimension Reduction Algorithm Based on Canonical Correlation Analysis [J]. Computer Engineering, 2020, 46(2): 1-10.
[8]	LIAO Hanyue, ZENG Jianping, WU Chengrong. A Model for Online Forum Traffic Prediction Integrated with Multiple Models [J]. Computer Engineering, 2020, 46(12): 60-66,72.
[9]	XIE Kun, RONG Yutian, HU Fengping, CHEN Huan, YAO Xiaolong. Random Forest Algorithm Based on Data Integration [J]. Computer Engineering, 2020, 46(12): 290-298.
[10]	WANG Liang, WANG Min, WANG Xiaopeng, LUO Wei, FENG Yu. Research on Network Queuing Delay Prediction Technology Based on Flow Calculation [J]. Computer Engineering, 2020, 46(10): 289-293,300.
[11]	DENG Yujing, WU Zhihao, LIN Youfang. Flight Passenger Load Factors Prediction Based on RNN Using Multi-Granularity Time Attention [J]. Computer Engineering, 2020, 46(1): 294-301.
[12]	WANG Huijian, LIU Zheng, LI Yun, LI Tao. Trend Prediction Method of Time Series Trends Based on Neural Network Language Model [J]. Computer Engineering, 2019, 45(7): 13-19,25.
[13]	LIU Zhi-Gang, DU Juan, HU Shao-Hua, LI Fen-Che. Time Series Prediction Method Based on Process Neural Networks [J]. Computer Engineering, 2012, 38(5): 199-201.
[14]	XUE Hai-dong; ZHU Qun-xiong. Time Series Prediction Algorithm Based on Structured Analogy [J]. Computer Engineering, 2010, 36(1): 211-214.

Please choose a citation manager

Content to export