A Riemannian Primal-dual Algorithm Based on Proximal Operator and its Application in Metric Learning.
Shijun WangBaocheng ZhuLintao MaYuan QiPublished in: IJCNN (2019)
Keyphrases
- metric learning
- distance metric
- semi supervised
- distance metric learning
- learning tasks
- kernel matrix
- pairwise
- machine learning and pattern recognition
- multi task
- mahalanobis distance
- dimensionality reduction
- distance function
- manifold learning
- feature space
- geodesic distance
- data sets
- linear transformation
- mahalanobis metric
- computer vision
- vector space
- euclidean distance
- semi supervised learning
- active learning
- labeled data
- unsupervised learning
- image classification
- database systems
- subject to linear constraints