Y
Yin Zhang
Researcher at Carnegie Mellon University
Publications - 14
Citations - 1852
Yin Zhang is an academic researcher from Carnegie Mellon University. The author has contributed to research in topics: Curse of dimensionality & Usability. The author has an hindex of 9, co-authored 12 publications receiving 1381 citations. Previous affiliations of Yin Zhang include Nanjing University.
Papers
More filters
Journal ArticleDOI
Understanding bag-of-words model: A statistical framework
Yin Zhang,Rong Jin,Zhi-Hua Zhou +2 more
TL;DR: A statistical framework which generalizes the bag-of-words representation, in which the visual words are generated by a statistical process rather than using a clustering algorithm, while the empirical performance is competitive to clustering-based method.
Journal ArticleDOI
Multilabel dimensionality reduction via dependence maximization
Yin Zhang,Zhi-Hua Zhou +1 more
TL;DR: Zhang et al. as mentioned in this paper proposed a multilabel dimensionality reduction method, MDDM, with two kinds of projection strategies, attempting to project the original data into a lower-dimensional feature space maximizing the dependence between the original feature description and the associated class labels.
Proceedings Article
Multi-label learning with weak label
Yuyin Sun,Yin Zhang,Zhi-Hua Zhou +2 more
TL;DR: The WELL (WEak Label Learning) method is proposed, which considers that the classification boundary for each label should go across low density regions, and that each label generally has much smaller number of positive examples than negative examples.
Journal ArticleDOI
Cost-Sensitive Face Recognition
Yin Zhang,Zhi-Hua Zhou +1 more
TL;DR: A framework is proposed which formulates the face recognition problem as a multiclass cost-sensitive learning task, and two theoretically sound methods for this task are developed.
Proceedings Article
Multi-label dimensionality reduction via dependence maximization
Yin Zhang,Zhi-Hua Zhou +1 more
TL;DR: This article proposes a multilabel dimensionality reduction method, MDDM, with two kinds of projection strategies, attempting to project the original data into a lower-dimensional feature space maximizing the dependence between the original feature description and the associated class labels.