Please use this identifier to cite or link to this item:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/80599Full metadata record
| ???org.dspace.app.webui.jsptag.ItemTag.dcfield??? | Value | Language |
|---|---|---|
| dc.contributor.advisor | 林守德(Shou-De Lin) | |
| dc.contributor.author | Chih-Chun Yang | en |
| dc.contributor.author | 楊之郡 | zh_TW |
| dc.date.accessioned | 2022-11-24T03:10:23Z | - |
| dc.date.available | 2021-11-03 | |
| dc.date.available | 2022-11-24T03:10:23Z | - |
| dc.date.copyright | 2021-11-03 | |
| dc.date.issued | 2021 | |
| dc.date.submitted | 2021-10-27 | |
| dc.identifier.citation | [1] S. Buuren and C. GroothuisOudshoorn. Mice: Multivariate imputation by chained equations in r. Journal of Statistical Software, 45, 12 2011. [2] D. Dua and C. Graff. UCI machine learning repository, 2017. [3] I. Goodfellow, J. PougetAbadie, M. Mirza, B. Xu, D. WardeFarley, S. Ozair, A. Courville, and Y. Bengio. Generative adversarial nets. 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. [4] O. Troyanskaya, M. Cantor, G. Sherlock, P. Brown, T. Hastie, R. Tibshirani, D. Bot stein, and R. B. Altman. Missing value estimation methods for DNA microarrays . Bioinformatics, 17(6):520–525, 06 2001. [5] A.Vaswani,N.Shazeer,N.Parmar,J.Uszkoreit,L.Jones,A.N.Gomez,L.u.Kaiser, and I. Polosukhin. Attention is all you need. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, editors, Advances in Neural Information Processing Systems, volume 30. Curran Associates, Inc., 2017. [6] J. Yoon, J. Jordon, and M. van der Schaar. GAIN: Missing data imputation using generative adversarial nets. In J. Dy and A. Krause, editors, Proceedings of the 35th International Conference on Machine Learning, volume 80 of Proceedings of Machine Learning Research, pages 5689–5698. PMLR, 10–15 Jul 2018. [7] J. You, X. Ma, Y. Ding, M. J. Kochenderfer, and J. Leskovec. Handling missing data with graph representation learning. In H. Larochelle, M. Ranzato, R. Hadsell, M. F. Balcan, and H. Lin, editors, Advances in Neural Information Processing Systems, volume 33, pages 19075–19087. Curran Associates, Inc., 2020. | |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/80599 | - |
| dc.description.abstract | 如何處理測試資料中的缺失值一直是一項備受關注的研究主題。對於學習在完整資料上且沒有做出預防措施的下游模型，我們必須先針對資料進行缺失值的填補才可進行預測。因此，如何學習一個良好的缺失值填補方法以及利用已知下游模型所提供的資訊將會是這項研究主題的關鍵。本論文旨在探討如何運用已知且不可更動的下游分類模型針對有缺失值的表格式資料進行預測。我們提出一項全新的架構，它包含了一個運用自注意力架構的缺失值填補模型以及一項迭代的下游任務標籤估計方法。架構中的缺失值填補模型可以替換成任何已知的模型。透過兩項自監督的學習任務以及從下游模型反饋的資訊，我們的缺失值填補模型在所有的競爭者中達到最佳的效果。我們透過大量的實驗來驗證我們提出的架構不只對於我們所提出的自注意力缺失值填補模型有所幫助，對於所有已知的缺失值填補方法也都可以提供進一步的效能提升。 | zh_TW |
| dc.description.provenance | Made available in DSpace on 2022-11-24T03:10:23Z (GMT). No. of bitstreams: 1 U0001-2210202118454100.pdf: 1290056 bytes, checksum: 99f8015c8cb0a5503f3f11a0f70415c0 (MD5) Previous issue date: 2021 | en |
| dc.description.tableofcontents | 口試委員審定書 P.i 摘要 P. iii Abstract P. iiii Contents P. vi List of Figures P. viii List of Tables P. xi 1 Introduction P. 1 2 Related Work P. 4 2.1 Missing Data Imputation P. 4 3 Methodology P. 6 3.1 Notations and Problem Formulation P. 6 3.2 Downstream-Aware Imputation FrameworkP. 6 3.3 Learning Imputation Model with SelfSupervision P. 7 3.3.1 Masking and Prediction Task P. 9 3.3.2 Missing Adversarial Task P. 10 3.3.3 Learning Objectives of Self-Supervised Tasks for Imputer P. 12 3.4 Exploiting Knowledge from the Prediction of Downstream Model P. 12 3.4.1 Iterative Downstream Label Estimation Algorithm P. 13 3.5 Full Optimization Pipeline of Our Framework P. 14 3.5.1 Full Learning Objectives for Imputer P. 15 3.5.2 Full Training Pipeline P. 15 4. Experiments P. 17 4.1 ExperimentSetup P. 17 4.1.1 Datasets P. 18 4.1.2 Baseline Methods P. 18 4.2 Comparison of Imputation Model P. 19 4.3 Analysis on Downstream Label Estimation Algorithm P. 20 4.4 Ablation Studies P. 21 5. Conclusion P. 23 Reference P. 24 | |
| dc.language.iso | zh-TW | |
| dc.subject | 深度學習 | zh_TW |
| dc.subject | 表格式資料 | zh_TW |
| dc.subject | 預測資料之缺失值填補 | zh_TW |
| dc.subject | 自監督學習 | zh_TW |
| dc.subject | 自注意力機制 | zh_TW |
| dc.subject | Self-Attention | en |
| dc.subject | Self-Supervised Learning | en |
| dc.subject | Missing Testing Data Imputation | en |
| dc.subject | Tabular Data | en |
| dc.subject | Deep Learning | en |
| dc.title | 基於自監督及迭代標籤估計之預測時資料缺失值處理 | zh_TW |
| dc.title | Handling Missing Data during Prediction based onSelfSupervision and Iterative Label Estimation | en |
| dc.date.schoolyear | 109-2 | |
| dc.description.degree | 碩士 | |
| dc.contributor.oralexamcommittee | 李政德(Hsin-Tsai Liu),駱宏毅(Chih-Yang Tseng) | |
| dc.subject.keyword | 深度學習,表格式資料,預測資料之缺失值填補,自監督學習,自注意力機制, | zh_TW |
| dc.subject.keyword | Deep Learning,Tabular Data,Missing Testing Data Imputation,Self-Supervised Learning,Self-Attention, | en |
| dc.relation.page | 25 | |
| dc.identifier.doi | 10.6342/NTU202104052 | |
| dc.rights.note | 同意授權(限校園內公開) | |
| dc.date.accepted | 2021-10-28 | |
| dc.contributor.author-college | 電機資訊學院 | zh_TW |
| dc.contributor.author-dept | 資訊工程學研究所 | zh_TW |
| Appears in Collections: | 資訊工程學系 | |
Files in This Item:
| File | Size | Format | |
|---|---|---|---|
| U0001-2210202118454100.pdf Access limited in NTU ip range | 1.26 MB | Adobe PDF |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.