Loading...
|
Please use this identifier to cite or link to this item:
https://nccur.lib.nccu.edu.tw/handle/140.119/153146
|
Title: | 分佈偏移與有限樣本下的最佳物料採購 Optimal Commodity Sourcing under Distributional Shifts and Limited Samples |
Authors: | 張聚洋 Chang, Chu-Yang |
Contributors: | 莊皓鈞 周彥君 Chuang, Hao-Chun Chou, Yen-Chun 張聚洋 Chang, Chu-Yang |
Keywords: | 分佈偏移 有限樣本 生成對抗模型 拔靴法 Distributional Shift Limited Sample GAN Bootstrap |
Date: | 2024 |
Issue Date: | 2024-09-04 14:02:54 (UTC+8) |
Abstract: | 本研究旨在探討生成對抗網絡(Generative Adversarial Network, GAN)和重採樣技術(Resampling)在原物料商品預測中的應用。通過引入這些技術,我們成功地提升了模型的預測性能,並為未來研究提供了寶貴的見解和方法論支持。
我們首先發現,GAN的鑑別層在解決數據分佈偏移(Distribution Shift)問題上表現出色。即使在數據不平衡和存在異常值的情況下,鑑別層依然能夠增強模型的穩定性和準確性。具體而言,引入鑑別層後,多數商品的預測誤差顯著降低。
重採樣技術在緩解時序數據量不足(Limited Sample)問題上也發揮了重要作用。通過使用移動區塊拔靴法(Moving Block Bootstrap),我們生成了多個重採樣數據集。這不僅增加了訓練數據的多樣性,還顯著改善了模型的泛化能力。實驗結果顯示,重採樣技術在多數商品上的應用效果顯著,尤其在銀和大豆的預測中,模型性能有了顯著提升。 This study aims to explore the application of Generative Adversarial Networks (GAN) and Resampling techniques in forecasting commodity prices. By incorporating these techniques, we have successfully enhanced the predictive performance of the model and provided valuable insights and methodological support for future research.
Firstly, we found that the discriminator layer in GAN performs excellently in addressing the problem of distribution shift. Even in cases of data imbalance and outliers, the discriminator layer can still enhance the model's stability and accuracy. Specifically, after introducing the discriminator layer, the prediction error for most commodities significantly decreased.
The resampling technique also plays an essential role in alleviating the problem of limited time series data. By using the Moving Block Bootstrap method, we generated multiple resampled datasets. This not only increased the diversity of the training data but also significantly improved the model's generalization capability. Experimental results show that the application of resampling techniques is significantly effective for most commodities, especially in the predictions of silver and soybeans, where the model's performance has significantly improved. |
Reference: | 1. Ben Ameur, H., Boubaker, S., Ftiti, Z., Louhichi, W., & Tissaoui, K. (2023). Forecasting commodity prices: Empirical evidence using deep learning tools. Annals of Operations Research, 1–19. 2. Bergmeir, C., Hyndman, R. J., & Benítez, J. M. (2016). Bagging exponential smoothing methods using stl decomposition and box–cox transformation. International Journal of Forecasting, 32(2), 303–312. https://doi.org/10.1016/j.ijforecast.2015.07.002 3. Busch, N., Crönert, T., Minner, S., Rettinger, M., & Sel, B. (2023). Deep learning for commodity procurement: Nonlinear data-driven optimization of hedging decisions. INFORMS Journal on Optimization, 5(3), 273–294. 4. Dong, K., Sun, R., & Dong, X. (2018). CO2 emissions, natural gas and renewables, economic growth: Assessing the evidence from China. Science of the Total Environment, 640, 293–302. 5. Dooley, G., & Lenihan, H. (2005). An assessment of time series methods in metal price forecasting. Resources Policy, 30(3), 208–217. 6. Fattahi, M. (2021). Resilient procurement planning for supply chains: A case study for sourcing a critical mineral material. Resources Policy, 74, 101093. 7. Hu, Z., Zhu, J., & Tse, K. (2013). Stocks market prediction using support vector machine. 2013 6th International Conference on Information Management, Innovation Management and Industrial Engineering, 2, 115–118. 8. Karasu, S., & Altan, A. (2022). Crude oil time series prediction model based on LSTM network with chaotic Henry gas solubility optimization. Energy, 242, 122964. 9. Khaidem, L., Saha, S., & Dey, S. R. (2016). Predicting the direction of stock market prices using random forest. arXiv preprint arXiv:1605.00003. 10. Kostrzewski, M., & Kostrzewska, J. (2019). Probabilistic electricity price forecasting with Bayesian stochastic volatility models. Energy Economics, 80, 610–620. https://doi.org/10.1016/j.eneco.2019.02.004 11. Mandl, C., & Minner, S. (2023). Data-driven optimization for commodity procurement under price uncertainty. Manufacturing & Service Operations Management, 25(2), 371–390. 12. Senoner, J., Chisari, F., Cherkaoui, R., & Kamwa, I. (2023). Addressing distributional shifts in operations management: The case of order fulfillment in customized production. Production and Operations Management, 32(10), 3022–3042. https://doi.org/10.1111/poms.14021 13. Shen, J., Qu, Y., Zhang, W., Yu, Y., McIlraith, S., & Weinberger, K. (2018). Proceedings of the thirty-second AAAI Conference on Artificial Intelligence (AAAI-18), the 30th Innovative Applications of Artificial Intelligence (IAAI-18), and the 8th AAAI Symposium on Educational Advances in Artificial Intelligence (EAAI-18). 14. Shrikumar, A., Greenside, P., Kundaje, A., Doina, P., & Yee, W. (2017). Proceedings of the 34th International Conference on Machine Learning. 15. Suryaningrat, I. B. (2016). Raw material procurement on agroindustrial supply chain management: A case survey of fruit processing industries in Indonesia. Agriculture and Agricultural Science Procedia, 9, 253–257. 16. Szarek, D., Bielak, Ł., & Wyłomańska, A. (2020). Long-term prediction of the metals’ prices using non-Gaussian time-inhomogeneous stochastic process. Physica A: Statistical Mechanics and its Applications, 555, 124659. 17. Taylor, S. J., & Letham, B. (2018). Forecasting at scale. The American Statistician, 72(1), 37–45. 18. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, Ł., & Polosukhin, I. (2017). Attention is all you need. Advances in Neural Information Processing Systems. |
Description: | 碩士 國立政治大學 資訊管理學系 111356006 |
Source URI: | http://thesis.lib.nccu.edu.tw/record/#G0111356006 |
Data Type: | thesis |
Appears in Collections: | [資訊管理學系] 學位論文
|
Files in This Item:
File |
Description |
Size | Format | |
600601.pdf | | 2921Kb | Adobe PDF | 0 | View/Open |
|
All items in 政大典藏 are protected by copyright, with all rights reserved.
|