Metric Learning from Relative Comparisons by Minimizing Squared Residual.
Eric Yi LiuZhishan GuoXiang ZhangVladimir JojicWei WangPublished in: ICDM (2012)
Keyphrases
- maximum variance unfolding
- metric learning
- subject to linear constraints
- distance metric
- machine learning and pattern recognition
- distance metric learning
- semi supervised
- pairwise
- multi task
- person re identification
- dimensionality reduction
- semi supervised clustering
- learning tasks
- feature space
- distance function
- euclidean distance
- semi supervised learning
- nearest neighbor classification
- margin maximization
- mahalanobis metric
- pattern recognition
- neural network
- high dimensional data
- labeled data
- least squares
- active learning
- knowledge base
- machine learning