由於數據易於獲得，時空模型變得更加重要，而圖卷積網路(Graph Convolution Network, GCN) 因為可以考慮圖中的信息，也可以考慮到它的鄰居為模型提供更多信息。然而，圖卷積網路GCN 卻很少應用於序列數據。 Attention-adjusted Graph Spatio-Temporal Network (AGSTN) 解決了圖卷積網路預測需要提前定義圖像並針對特定類型數據的問題，結合多層圖卷積網路(Multi-Graph Convolution Network, MGCN)和注意力調整(attention)來學習隨時間變化的空間和時間相關性，最終順利將 圖卷積網路應用於時間序列預測。
本文基於AGSTN的方法，並將此模型應用於一般數據集，除了希望解決原始序列模型(Long Short-Term Memory, LSTM)在預測中遇到的變量相關問題，我們還通過數據擴增(Data Augmentation)和正則化(Regularization)的方法解決了數據稀缺帶來的收斂問題，從原本的約50%的收斂機率進步到100%收斂。最後，將修改後的模型用於製造業相關數據，並使用不同的時間區段進行模型驗證。此外，在修正模型中我們還添加了時間序列分解特徵和防疫策略的變量等附加特徵來替換原來的本徵模式函數(Intrinsic Mode Function, IMF) 特徵，兩者都能改善模型表現，前者可將評估模型的均方誤差(Mean Squared Error, MSE)改善約20%；而後者則可使預測變量在變動幅度較大的時間範圍內預測較為準確。

Spatial-Temporal Model becomes more important due to the easy availability of data, and the GCN can provide more information to the model because it can take into account the information in the graph but also to its neighbors. However, GCN is rarely applied to sequence data. Attention-adjusted Graph Spatio-Temporal Network (AGSTN) solves the problem that GCN prediction needs to define images in advance and target specific types of data. Combining multi-graph convolution and attention adjustment to learn spatial and temporal correlations over time. Then GCN is applied to time series forecasting.
This paper is based on AGSTN, and this model is applied to non-sensor datasets, we hope to solve the variable-related problems encountered in the prediction of the original sequence model. At the same time, we also solves the convergence problem caused by scarcity of data through a method of data Augmentation and regularization. Convergence probability is improved from 50% to 100%.
Finally, the modified model was used in manufacturing-related data, and use different time periods to do model verification. Furthermore, additional features such as time series decomposition features and policy variable were added to replace the original IMF features, both of them make the model perform better. The mean squared error (MSE) of the model evaluation is improved about 20% in time series decomposition features; the latter makes the prediction to be more accurate in time horizons with large fluctuations.

Bansal, M. A., Sharma, D. R., & Kathuria, D. M. (2021). A Systematic Review on Data Scarcity Problem in Deep Learning: Solution and Applications. ACM Computing Surveys (CSUR).
Box, G. E., Jenkins, G. M., Reinsel, G. C., & Ljung, G. M. (2015). Time series analysis: forecasting and control: John Wiley & Sons.
Breiman, L. (1996a). Bagging predictors. Machine learning, 24(2), 123-140.
Breiman, L. (1996b). Stacked regressions. Machine learning, 24(1), 49-64.
Cai, M., & Luo, J. (2020). Influence of COVID-19 on manufacturing industry and corresponding countermeasures from supply chain perspective. Journal of Shanghai Jiaotong University (Science), 25(4), 409-416.
Campos, D., Kieu, T., Guo, C., Huang, F., Zheng, K., Yang, B., & Jensen, C. S. (2021). Unsupervised Time Series Outlier Detection with Diversity-Driven Convolutional Ensembles--Extended Version. arXiv preprint arXiv:2111.11108.
Chen, A.-S., Leung, M. T., & Daouk, H. (2003). Application of neural networks to an emerging financial market: forecasting and trading the Taiwan Stock Index. Computers & Operations Research, 30(6), 901-923.
Chiaradonna, S., Di Giandomenico, F., & Murru, N. (2020). On enhancing efficiency and accuracy of particle swarm optimization algorithms. arXiv preprint arXiv:2004.06326.
Chung, J., Gulcehre, C., Cho, K., & Bengio, Y. (2014). Empirical evaluation of gated recurrent neural networks on sequence modeling. arXiv preprint arXiv:1412.3555.
Cirstea, R.-G., Kieu, T., Guo, C., Yang, B., & Pan, S. J. (2021). EnhanceNet: Plugin neural networks for enhancing correlated time series forecasting. Paper presented at the 2021 IEEE 37th International Conference on Data Engineering (ICDE).
Cirstea, R.-G., Micu, D.-V., Muresan, G.-M., Guo, C., & Yang, B. (2018). Correlated time series forecasting using multi-task deep neural networks. Paper presented at the Proceedings of the 27th acm international conference on information and knowledge management.
Cirstea, R.-G., Yang, B., & Guo, C. (2019). Graph attention recurrent neural networks for correlated time series forecasting. MileTS19@ KDD.
del Rio-Chanona, R. M., Mealy, P., Pichler, A., Lafond, F., & Farmer, J. D. (2020). Supply and demand shocks in the COVID-19 pandemic: An industry and occupation perspective. Oxford Review of Economic Policy, 36(Supplement_1), S94-S137.
Diao, Z., Wang, X., Zhang, D., Liu, Y., Xie, K., & He, S. (2019). Dynamic spatial-temporal graph convolutional neural networks for traffic forecasting. Paper presented at the Proceedings of the AAAI conference on artificial intelligence.
Effrosynidis, D., & Arampatzis, A. (2021). An evaluation of feature selection methods for environmental data. Ecological Informatics, 61, 101224.
Geng, X., Li, Y., Wang, L., Zhang, L., Yang, Q., Ye, J., & Liu, Y. (2019). Spatiotemporal multi-graph convolution network for ride-hailing demand forecasting. Paper presented at the Proceedings of the AAAI conference on artificial intelligence.
Guo, S., Lin, Y., Feng, N., Song, C., & Wan, H. (2019). Attention based spatial-temporal graph convolutional networks for traffic flow forecasting. Paper presented at the Proceedings of the AAAI conference on artificial intelligence.
Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural computation, 9(8), 1735-1780.
Hyndman, R., & Athanasopoulos, G. (2012). 6.1 Time series components. Forecasting: Principles and Practice.
Laine, S., & Aila, T. (2016). Temporal ensembling for semi-supervised learning. arXiv preprint arXiv:1610.02242.
Lee, C.-Y., & Chien, C.-F. (2014). Stochastic programming for vendor portfolio selection and order allocation under delivery uncertainty. OR spectrum, 36(3), 761-797.
Lu, Y.-J., & Li, C.-T. (2020). AGSTN: Learning Attention-adjusted Graph Spatio-Temporal Networks for Short-term Urban Sensor Value Forecasting. Paper presented at the 2020 IEEE International Conference on Data Mining (ICDM).
Maaliw, R. R., Ballera, M. A., Mabunga, Z. P., Mahusay, A. T., Dejelo, D. A., & Seño, M. P. (2021). An ensemble machine learning approach for time series forecasting of COVID-19 cases. Paper presented at the 2021 IEEE 12th Annual Information Technology, Electronics and Mobile Communication Conference (IEMCON).
McKenzie, E. (1984). General exponential smoothing and the equivalent ARMA process. Journal of Forecasting, 3(3), 333-344.
Meredith, J., & Akinc, U. (2007). Characterizing and structuring a new make-to-forecast production strategy. Journal of Operations Management, 25(3), 623-642.
Miyato, T., Maeda, S.-i., Koyama, M., & Ishii, S. (2018). Virtual adversarial training: a regularization method for supervised and semi-supervised learning. IEEE transactions on pattern analysis and machine intelligence, 41(8), 1979-1993.
Moreno-Barea, F. J., Strazzera, F., Jerez, J. M., Urda, D., & Franco, L. (2018). Forward noise adjustment scheme for data augmentation. Paper presented at the 2018 IEEE symposium series on computational intelligence (SSCI).
Oord, A. v. d., Dieleman, S., Zen, H., Simonyan, K., Vinyals, O., Graves, A., . . . Kavukcuoglu, K. (2016). Wavenet: A generative model for raw audio. arXiv preprint arXiv:1609.03499.
Opitz, D., & Maclin, R. (1999). Popular ensemble methods: An empirical study. Journal of artificial intelligence research, 11, 169-198.
Polikar, R. (2006). Ensemble based systems in decision making. IEEE Circuits and systems magazine, 6(3), 21-45.
Qian, Y., Hu, H., & Tan, T. (2019). Data augmentation using generative adversarial networks for robust speech recognition. Speech Communication, 114, 1-9.
Rong, Y., Huang, W., Xu, T., & Huang, J. (2019). Dropedge: Towards deep graph convolutional networks on node classification. arXiv preprint arXiv:1907.10903.
Sajjadi, M., Javanmardi, M., & Tasdizen, T. (2016). Regularization with stochastic transformations and perturbations for deep semi-supervised learning. Advances in neural information processing systems, 29.
Smyth, P., & Wolpert, D. (1999). Linearly combining density estimators via stacking. Machine learning, 36(1), 59-83.
Sukanya, C. (2020). A prediction of operation and supply chain management in post covid 19 era: in the case manufacturers in Thailand.
Telles, C. R., Roy, A., Ajmal, M. R., Mustafa, S. K., Ahmad, M. A., de la Serna, J. M., . . . Rosales, M. H. (2021). The impact of COVID-19 management policies tailored to airborne SARS-CoV-2 transmission: Policy analysis. JMIR Public Health and Surveillance, 7(4), e20699.
Tseng, V. S., Jia-Ching Ying, J., Wong, S. T. C., Cook, D. J., & Liu, J. (2020). Computational Intelligence Techniques for Combating COVID-19: A Survey. IEEE Computational Intelligence Magazine, 15(4), 10-22. doi:10.1109/mci.2020.3019873
Wei, G., Zhao, J., Feng, Y., He, A., & Yu, J. (2020). A novel hybrid feature selection method based on dynamic feature importance. Applied Soft Computing, 93, 106337.
Witten, I. H., Frank, E., Hall, M. A., Pal, C. J., & DATA, M. (2005). Practical machine learning tools and techniques. Paper presented at the Data Mining.
Wolpert, D. H. (1992). Stacked generalization. Neural networks, 5(2), 241-259.
Wu, R., Yan, S., Shan, Y., Dang, Q., & Sun, G. (2015). Deep image: Scaling up image recognition. arXiv preprint arXiv:1501.02876, 7(8).
Wu, Y., & Tan, H. (2016). Short-term traffic flow forecasting with spatial-temporal correlation in a hybrid deep learning framework. arXiv preprint arXiv:1612.01022.
Yousaf, M., Zahir, S., Riaz, M., Hussain, S. M., & Shah, K. (2020). Statistical analysis of forecasting COVID-19 for upcoming month in Pakistan. Chaos, Solitons & Fractals, 138, 109926.
Zhang, C., Liu, M., Liu, Z., Yang, C., Zhang, L., & Han, J. (2018). Spatiotemporal activity modeling under data scarcity: A graph-regularized cross-modal embedding approach. Paper presented at the Proceedings of the AAAI Conference on Artificial Intelligence.
Zhang, J., Zheng, Y., Qi, D., Li, R., & Yi, X. (2016). DNN-based prediction model for spatio-temporal data. Paper presented at the Proceedings of the 24th ACM SIGSPATIAL international conference on advances in geographic information systems.
Zheng, C., Zong, B., Cheng, W., Song, D., Ni, J., Yu, W., . . . Wang, W. (2020). Robust graph representation learning via neural sparsification. Paper presented at the International Conference on Machine Learning.
吳宗諭. (2021). 資料科學與條件風險價值強化學習於原油價格預測與採購決策