跳到主要內容

臺灣博碩士論文加值系統

(3.236.84.188) 您好!臺灣時間:2021/08/06 12:48
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

我願授權國圖
: 
twitterline
研究生:尤聖棨
研究生(外文):Sheng-Chi You
論文名稱:在觀念偏移資料中運用動態抑制更新改善線上學習
論文名稱(外文):Dynamic Unlearning for Online learning on Concept-drifting Data
指導教授:林軒田
指導教授(外文):Hsuan-Tien Lin
口試委員:林守德李育杰
口試委員(外文):Shou-De LinYuh-Jye Lee
口試日期:2015-06-30
學位類別:碩士
校院名稱:國立臺灣大學
系所名稱:資訊工程學研究所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2015
畢業學年度:103
語文別:英文
論文頁數:27
中文關鍵詞:線上學習觀念偏移資料
外文關鍵詞:Online learningconcpet drift
相關次數:
  • 被引用被引用:0
  • 點閱點閱:136
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
現實生活中的線上學習問題,通常伴隨著隨時改變的學習目標函式。這種改變「觀念偏移」,通常會降低線上學習演算法的效能。因此許多研究針對此嘗試分析資料的統計特性、提出偵測偏移改變的門檻機制。或者直接應用框格窗的方法維護較新的資料集合,反應出最近資料的觀念偏移。然而,鮮少研究同時專注於「偵測」偏移並維護「部份」資料來改善學習演算法。我們提出了一個基於現有的線上學習演算法的框架,在觀念偏移的資料提昇效能。此框架利用檢查抑制更新資料點(反更新)來檢查是否為觀念偏移,讓原算法得到更好的學習效果。此框架會根據套用的學習算法和處理的資料特性,讓學習演算法自然地產生動態框格窗來維持部份資料以對應資料觀念偏移的走勢。我們提出數種不同的抑制更新策略,並套用在三個經典得線上學習演算法。實驗結果顯示此框架在不同的觀念偏移資料下能有效提昇原先的學習演算法。

Real-world online learning applications often face data coming from changing target functions. Such changes, called the concept drift, degrade the performance of traditional online learning algorithms. Thus, many existing works focus on detecting concept drift based on statistical evidence. Other works use sliding window or similar mechanisms to select the data that closely reflect current concept. Nevertheless, few works study how the detection and selection techniques can be combined to improve the learning performance. We propose a framework on top of existing online learning algorithms to improve the learning performance under concept drift. The framework detects the possible concept drift by checking whether forgetting some older data may be helpful, and then conduct forgetting through a step called unlearning. The framework effectively results in a dynamic sliding window that selects the data flexibly for different kinds of concept drifts. We design concrete approaches from the framework based on three popular online learning algorithms. Empirical results show that the framework consistently improves those online learning algorithms on ten synthetic data sets and two real-world data sets.

誌謝 ii
中文摘要 iii
Abstract iv
1 Introduction 1
2 Preliminaries 3
2.1 Online Learning 3
2.2 Concept Drift 5
2.3 PA-based Online Learning Algorithms 8
3 Unlearning Framework 11
3.1 Greedy Unlearning Concept 11
3.2 Unlearning Decision Function 12
3.2.1 Hinge-Loss condition 12
3.2.2 Regularized term and confidence 13
3.3 Target for Unlearning test 14
3.3.1 Forwarding-removing 14
3.3.2 Stack-removing 15
3.3.3 Queue-removing 16
3.3.4 Selecting-removing 16
4 Empirical Evaluation 18
4.1 Data and Experimental setup 18
4.2 Results and Discussion 19
5 Conclusion 25
Bibliography 26

[1] M. Baena-García, J. del Campo-Ávila, R. Fidalgo, A. Bifet, R. Gavaldà, and R. Morales-Bueno. Early drift detection method. 2006.
[2] A. Bifet and R. Gavalda. Learning from time-changing data with adaptive windowing. 2007.
[3] K. Crammer, O. Dekel, J. Keshet, S. Shalev-Shwartz, and Y. Singer. Online passiveaggressive algorithms. The Journal of Machine Learning Research, 7:551–585, 2006.
[4] K. Crammer, A. Kulesza, and M. Dredze. Adaptive regularization of weight vectors. Advances in neural information processing systems, pages 414–422, 2009.
[5] M. Dredze, K. Crammer, and F. Pereira. Confidence-weighted linear classification. Proceedings of the 25th international conference on Machine learning, pages 264–271, 2008.
[6] J. Gama, P. Medas, G. Castillo, and P. Rodrigues. Learning with drift detection. Advances in Artificial Intelligence–SBIA 2004, pages 286–295, 2004.
[7] J. Gama, I. Žliobaitė, A. Bifet, M. Pechenizkiy, and A. Bouchachia. A survey on concept drift adaptation. ACM Computing Surveys (CSUR), 46(4):44, 2014.
[8] M. Harries and N. S. Wales. Splice-2 comparative evaluation: Electricity pricing. 1999.
[9] G. Hulten, L. Spencer, and P. Domingos. Mining time-changing data streams. Proceedings of the seventh ACM SIGKDD international conference on Knowledge discovery and data mining, pages 97–106, 2001.
[10] J. Z. Kolter and M. A. Maloof. Dynamic weighted majority: An ensemble method for drifting concepts. The Journal of Machine Learning Research, 8:2755–2790, 2007.
[11] Y. LeCun and C. Cortes. Mnist handwritten digit database. AT&T Labs [Online]. Available: http://yann. lecun. com/exdb/mnist, 2010.
[12] K. Nishida and K. Yamauchi. Detecting concept drift using statistical testing. Discovery Science, pages 264–269, 2007.
[13] K. Ralf and T. Joachims. Detecting concept drift with support vector machines. 2000.
[14] S. Shalev-Shwartz. Online learning and online convex optimization. Foundations and Trends in Machine Learning, 4(2):107–194, 2011.
[15] P. Sobhani and H. Beigy. New drift detection method for data streams. Adaptive and Intelligent Systems, page 88, 2011.
[16] Tsymbal and Alexey. The problem of concept drift: definitions and related work. Computer Science Department, Trinity College Dublin, 106, 2004.
[17] G. Widmer and M. Kubat. Learning in the presence of concept drift and hidden contexts. Machine learning, 23(1):69–101, 1996.


QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關論文
 
無相關期刊
 
無相關點閱論文