請用此 Handle URI 來引用此文件：
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/81360完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 陳孟彰(Meng-Chang Chen) | |
| dc.contributor.author | Hsin-Chih Yang | en |
| dc.contributor.author | 楊信之 | zh_TW |
| dc.date.accessioned | 2022-11-24T03:45:25Z | - |
| dc.date.available | 2022-02-21 | |
| dc.date.available | 2022-11-24T03:45:25Z | - |
| dc.date.copyright | 2022-02-21 | |
| dc.date.issued | 2022 | |
| dc.date.submitted | 2022-01-26 | |
| dc.identifier.citation | Hengxu Lin, Dong Zhou, Weiqing Liu, and Jiang Bian. Learning multiple stock trading patterns with temporal routing adaptor and optimal transport. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery amp; Data Mining, KDD ’21, page 1017–1026, New York, NY, USA, 2021. Association for Computing Machinery. ISBN 9781450383325. doi: 10.1145/3447548.3467358. URL https: //doi.org/10.1145/3447548.3467358. Yumo Xu and Shay B. Cohen. Stock movement prediction from tweets and historical prices. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1970–1979, Melbourne, Australia, July 2018. Association for Computational Linguistics. doi: 10.18653/v1/P181183. URL https://aclanthology.org/P181183. Ling Cai, Krzysztof Janowicz, Gengchen Mai, Bo Yan, and Rui Zhu. Traffic transformer: Capturing the continuity and periodicity of time series for traffic forecasting. Transactions in GIS, 24, 06 2020. doi: 10.1111/tgis.12644. Ilya Sutskever, Oriol Vinyals, and Quoc V Le. Sequence to sequence learning with neural networks. In Z. Ghahramani, M. Welling, C. Cortes, N. Lawrence, and K. Q. Weinberger, editors, Advances in Neural Information Processing Systems, volume 27. Curran Associates, Inc., 2014. URL https://proceedings.neurips.cc/paper/2014/file/ a14ac55a4f27472c5d894ec1c3c743d2Paper.pdf. Dario Jozinović, Anthony Lomax, Ivan Štajduhar, and Alberto Michelini. Rapid prediction of earthquake ground shaking intensity using raw waveform data and a convolutional neural network. Geophysical Journal International, 222(2):1379– 1389, 05 2020. ISSN 0956540X. doi: 10.1093/gji/ggaa233. URL https: //doi.org/10.1093/gji/ggaa233. Christian RequenaMesa, Vitus Benson, Markus Reichstein, Jakob Runge, and Joachim Denzler. Earthnet2021: A largescale dataset and challenge for earth surface forecasting as a guided video prediction task. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pages 1132–1142, 2021. Shuo Wang, Yanran Li, Jiang Zhang, Qingye Meng, Lingwei Meng, and Fei Gao. Pm2.5gnn: A domain knowledge enhanced graph neural network for pm2.5 forecasting. Proceedings of the 28th International Conference on Advances in Geographic Information Systems, 2020. MingChuan Yang and Meng Chang Chen. Composite neural network: Theory and application to pm2.5 prediction. IEEE Transactions on Knowledge and Data Engineering, pages 1–1, 2021. doi: 10.1109/TKDE.2021.3099135. Daizong Ding, Mi Zhang, Xudong Pan, Min Yang, and Xiangnan He. Modeling extreme events in time series prediction. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery amp; Data Mining, KDD ’19, page 1114–1122, New York, NY, USA, 2019. Association for Computing Machinery. ISBN 9781450362016. doi: 10.1145/3292500.3330896. URL https://doi.org/10.1145/3292500.3330896. Haibo He and Edwardo A. Garcia. Learning from imbalanced data. IEEE Transactions on Knowledge and Data Engineering, 21(9):1263–1284, 2009. doi: 10.1109/TKDE.2008.239. Nitesh V. Chawla, Nathalie Japkowicz, and Aleksander Kotcz. Editorial: Special issue on learning from imbalanced data sets. SIGKDD Explor. Newsl., 6(1):1– 6, jun 2004. ISSN 19310145. doi: 10.1145/1007730.1007733. URL https: //doi.org/10.1145/1007730.1007733. S. El Adlouni, B. Bobée, and T.B.M.J. Ouarda. On the tails of extreme event distributions in hydrology. Journal of Hydrology, 355(1):16–33, 2008. ISSN 0022 1694. doi: https://doi.org/10.1016/j.jhydrol.2008.02.011. URL https://www. sciencedirect.com/science/article/pii/S0022169408000887. Nikolay Pavlovich Laptev, Jason Yosinski, Li Erran Li, and Slawek Smyl. Timeseries extreme event forecasting with neural networks at uber. 2017 Yin Cui, Menglin Jia, TsungYi Lin, Yang Song, and Serge J. Belongie. Classbalanced loss based on effective number of samples. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 9260–9269, 2019. TsungYi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Dollar. Focal loss for dense object detection. In Proceedings of the IEEE International Conference on Computer Vision (ICCV), Oct 2017. Junyoung Chung, Çaglar Gülçehre, Kyunghyun Cho, and Yoshua Bengio. Empirical evaluation of gated recurrent neural networks on sequence modeling. ArXiv, abs/ 1412.3555, 2014. Sepp Hochreiter and Jürgen Schmidhuber. Long shortterm memory. Neural computation, 9(8):1735–1780, 1997. Ilya Sutskever, Oriol Vinyals, and Quoc V. Le. Sequence to sequence learning with neural networks. In NIPS, 2014. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Ł ukasz Kaiser, and Illia Polosukhin. Attention is all you need. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, editors, Advancesin Neural Information Processing Systems, volume 30. Curran Associates, Inc., 2017. URL https://proceedings.neurips.cc/paper/ 2017/file/3f5ee243547dee91fbd053c1c4a845aaPaper.pdf. Sergio Albeverio, Volker Jentsch, and Holger Kantz. Extreme events in nature and society. 2006. Sergey Foss, Dmitry Korshunov, and Stan Zachary. HeavyTailed and LongTailed Distributions, pages 7–38. Springer New York, New York, NY, 2011. ISBN 978 1441994738. doi: 10.1007/9781441994738_2. URL https://doi.org/ 10.1007/9781441994738_2. | |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/81360 | - |
| dc.description.abstract | 在測量空氣品質時，PM2.5的濃度是一個非常重要的指標。然而，在過往的時序資料領域的工作大多集中在預測短時間未來數值或長期的趨勢走向上，而沒有考慮到自然的時序資料所面臨的極端事件議題。這篇論文提出了三個基於解決極端資料和時序預測應用在PM2.5的損失函數和模型架構---極端事件損失函數、 Transformer 和 Composite Network 。首先，很少有研究在針對極端事件推導出相對應的損失函數。我們根據極端數值分佈結合非均衡資料的損失函數，得到的新穎目標函數。同時討論在不同目標函數下的訓練結果。第二，我們採用了 Transformer 的神經網路模型來取代過往時序資料中常見使用的循環神經網路。第三，為了解決極端事件造成模型難以訓練的問題，我們提出 Composite Network 的神經網路架構。與先前利用一個模型來預測不同，我們先把原始問題根據資料或損失函數拆解成更小的子問題並分別訓練，最後合併回歸原始問題，讓合併的模型可以針對不同的情況預測出更加準確的數值。我們的資料集採用臺灣中央氣象局所提供的2014到2020，總共七年的空氣品質資料。綜上所述，本研究為未來基於極端資料、非均衡資料和時序資料的預測引入了新的研究方向。 | zh_TW |
| dc.description.provenance | Made available in DSpace on 2022-11-24T03:45:25Z (GMT). No. of bitstreams: 1 U0001-2101202202451400.pdf: 1744909 bytes, checksum: 3d707a7918e0883ed6681c70c8d8b07c (MD5) Previous issue date: 2022 | en |
| dc.description.tableofcontents | Verification Letter from the Oral Examination Committee i Acknowledgements iii 摘要 v Abstract vii Contents ix List of Figures xi List of Tables xiii Chapter 1 Introduction 1 1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Overview of Related Works . . . . . . . . . . . . . . . . . . . . . . 3 1.3 Organization of This Thesis . . . . . . . . . . . . . . . . . . . . . . 3 Chapter 2 Preliminaries 5 2.1 Problem Formulation on PM2.5 . . . . . . . . . . . . . . . . . . . . 5 2.2 Extreme Value Theory . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.3 Extreme Event Problem in PM2.5 . . . . . . . . . . . . . . . . . . . 8 Chapter 3 Material and Method 15 3.1 Extreme Event Loss . . . . . . . . . . . . . . . . . . . . . . . . . . 15 3.2 Transformer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.3 Composite Network . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Chapter 4 Experiment and Results 25 4.1 Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 4.2 Influence of of Hyperparameters in Extreme Event . . . . . . . . . . 26 4.3 Comparison of different models . . . . . . . . . . . . . . . . . . . . 27 4.4 Prediction Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Chapter 5 Conclusion 33 References 35 | |
| dc.language.iso | en | |
| dc.subject | 非均衡資料 | zh_TW |
| dc.subject | 時序資料 | zh_TW |
| dc.subject | 重尾分布 | zh_TW |
| dc.subject | 極端事件 | zh_TW |
| dc.subject | Extreme Event | en |
| dc.subject | Heavy-tailed Distribution | en |
| dc.subject | Imbalanced Data | en |
| dc.subject | Time Series Data | en |
| dc.title | 基於PM2.5長時間序列深度學習的極端事件預測 | zh_TW |
| dc.title | Extreme event prediction based on PM2.5 long-term time series deep learning | en |
| dc.date.schoolyear | 110-1 | |
| dc.description.degree | 碩士 | |
| dc.contributor.coadvisor | 黃俊郎(Jiun-Lang Huang) | |
| dc.contributor.oralexamcommittee | 陳伶志(Te-Hsin Yang),楊名全(Chen-Jung Lin),(Min-Jay Chung) | |
| dc.subject.keyword | 極端事件,時序資料,非均衡資料,重尾分布, | zh_TW |
| dc.subject.keyword | Extreme Event,Time Series Data,Imbalanced Data,Heavy-tailed Distribution, | en |
| dc.relation.page | 38 | |
| dc.identifier.doi | 10.6342/NTU202200131 | |
| dc.rights.note | 同意授權(限校園內公開) | |
| dc.date.accepted | 2022-01-27 | |
| dc.contributor.author-college | 電機資訊學院 | zh_TW |
| dc.contributor.author-dept | 資料科學學位學程 | zh_TW |
| 顯示於系所單位： | 資料科學學位學程 | |
文件中的檔案：
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| U0001-2101202202451400.pdf 授權僅限NTU校內IP使用（校園外請利用VPN校外連線服務） | 1.7 MB | Adobe PDF |
系統中的文件，除了特別指名其著作權條款之外，均受到著作權保護，並且保留所有的權利。