Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Jun 27:109413.
doi: 10.1016/j.cie.2023.109413. Online ahead of print.

Graph Spatio-Temporal Networks for Manufacturing Sales Forecast and Prevention Policies in Pandemic Era

Chia-Yen Lee  1 , Shu-Huei Yang  2
Affiliations

Graph Spatio-Temporal Networks for Manufacturing Sales Forecast and Prevention Policies in Pandemic Era

Chia-Yen Lee et al. Comput Ind Eng. .

Abstract

Worldwide manufacturing industries are significantly affected by COVID-19 pandemic because of their production characteristics with low-cost country sourcing, globalization, and inventory level. To analyze the correlated time series, spatial-temporal model becomes more attractive, and the graph convolution network (GCN) is also commonly used to provide more information to the nodes and its neighbors in the graph. Recently, attention-adjusted graph spatio-temporal network (AGSTN) was proposed to address the problem of pre-defined graph in GCN by combining multi-graph convolution and attention adjustment to learn spatial and temporal correlations over time. However, AGSTN may show potential problem with limited small non-sensor data; particularly, convergence issue. This study proposes several variants of AGSTN and applies them to non-sensor data. We suggest data augmentation and regularization techniques such as edge selection, time series decomposition, prevention policies to improve AGSTN. An empirical study of worldwide manufacturing industries in pandemic era was conducted to validate the proposed variants. The results show that the proposed variants significantly improve the prediction performance at least around 20% on mean squared error (MSE) and convergence problem.

Keywords: Correlated time series; Data augmentation; Data scarcity; Graph convolutional networks; Regularization.

PubMed Disclaimer

Conflict of interest statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Figures

Figure 1
Figure 1
The architecture of AGSTN (Lu and Li, 2020)
Figure 2
Figure 2
An illustration of NeuralSparse (revised from Zheng et al., 2020)
Figure 3
Figure 3
The architecture of the AGSTN variants
Figure 4
Figure 4
The proposed research framework
Figure 5
Figure 5
Linear interpolation of data
Figure 6
Figure 6
Time series decomposition for new order
Figure 7
Figure 7
US new orders in each period of big events
Figure 8
Figure 8
Illustration of investigating the selected variables
Figure 9
Figure 9
Comparison between AGSTN and AGSTN+NS
Figure 10
Figure 10
Comparison between AGSTN+PP and AGSTN+NS+TSD
Figure 11
Figure 11
Model comparison in time period A
Figure 12
Figure 12
Model comparison in time period B
See this image and copyright information in PMC

References

    1. Bansal M.A., Sharma D.R., Kathuria D.M. A systematic review on data scarcity problem in deep learning: Solution and aapplications. ACM Computing Surveys. 2022;54(10s), article no.: 208:1–29.
    1. Cai M., Luo J. Influence of COVID-19 on manufacturing industry and corresponding countermeasures from supply chain perspective. Journal of Shanghai Jiaotong University (Science) 2020;25(4):409–416. - PMC - PubMed
    1. Campos D., Kieu T., Guo C., Huang F., Zheng K., Yang B., Jensen C.S. Unsupervised time series outlier detection with diversity-driven convolutional ensembles. Proceedings of the VLDB Endowment. 2021;15(3):611–623.
    1. Chung, J., Gulcehre, C., Cho, K., & Bengio, Y. (2014). Empirical evaluation of gated recurrent neural networks on sequence modeling. NIPS 2014 Deep Learning and Representation Learning Workshop.
    1. Cirstea, R.-G., Kieu, T., Guo, C., Yang, B., & Pan, S. J. (2021). EnhanceNet: Plugin neural networks for enhancing correlated time series forecasting. 2021 IEEE 37th International Conference on Data Engineering (ICDE).

LinkOut - more resources