Heteroscedastic max-min distance analysis

Abstract

Many discriminant analysis methods such as LDA and HLDA actually maximize the average pairwise distances between classes, which often causes the class separation problem. Max-min distance analysis (MMDA) addresses this problem by maximizing the minimum pairwise distance in the latent subspace, but it is developed under the homoscedastic assumption. This… (More)
DOI: 10.1109/CVPR.2015.7299084

Topics

1 Figure or Table

Statistics

050100150201520162017
Citations per Year

Citation Velocity: 34

Averaging 34 citations per year over the last 3 years.

Learn more about how we calculate this metric in our FAQ.

Cite this paper

@article{Su2015HeteroscedasticMD, title={Heteroscedastic max-min distance analysis}, author={Bing Su and Xiaoqing Ding and Changsong Liu and Ying Wu}, journal={2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, year={2015}, pages={4539-4547} }