| Summary: | 碩士 === 國立臺灣大學 === 資訊工程學研究所 === 100 === 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.
|
|---|
