透過您的圖書館登入
IP:3.147.61.3
  • 學位論文

多標籤分類問題中考量特徵的標籤空間降維法

Feature-aware Label Space Dimension Reduction for Multi-label Classification Problem

指導教授 : 林軒田
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


標籤空間降維法(Label Space Dimension Reduction)在多標籤分 類 問 題 (Multi-lable Classification Problem) 中 是 一 個 有 效 且 有 效 率 的 方 法 。 現 存 的 標 籤 空 間 降 維 法 ( 例 如 : 壓 縮 感 知 (Compressive Sensing)與主要標籤空間變換(Principal Label Space Transformation)) 只利用資料中標籤部份的資訊。 在本論文中,我們提出一個能夠同時 考量特徵與標籤之資訊的標籤空間降維法。 此稱為條件式主要標籤空 間變換(Conditional Principal Label Space Transformation)之演算法的 設計目的是最小化廣泛使用的漢明虧損的上界。 此方法的最小化步驟 能夠透過有效率的執行奇異值分解達成。 此外,此方法能夠擴展至核 函數方法,核函數方法允許使用更精密複雜的特徵組合去幫助標籤空 間降維。 實驗結果顯示此方法在標籤分類問題中確實比其他現存方法 更有效。

並列摘要


Label space dimension reduction (LSDR) is an efficient and effective paradigm for multi-label classification with many classes. Existing approaches to LSDR, such as compressive sensing and principal label space transformation, ex- ploit only the label part of the dataset, but not the feature part. In this thesis, we propose a novel approach to LSDR that considers both the label and the feature parts. The approach, called conditional principal label space trans- formation, is based on minimizing an upper bound of the popular Hamming loss. The minimization step of the approach can be carried out efficiently by a simple use of singular value decomposition. In addition, the approach can be extended to a kernelized version that allows the use of sophisticated feature combinations to assist LSDR. The experimental results verify that the proposed approach is more effective than existing ones to LSDR across many real-world datasets.

參考文獻


principal component analysis. A Practical Approach to Microarray Data Analysis,
Learning and Principles and Practice of Knowledge Discovery in Databases 2008
Discovery Challenge, 2008.
[2] M. R. Boutell, J. Luo, X. Shen, and C. M. Brown. Learning multi-label scene clas-
sification. Pattern Recognition, 37:1757–1771, 2004.

延伸閱讀