Metric Learning in Dissimilarity Space for Improved Nearest Neighbor Performance.
Robert P. W. DuinManuele BicegoMauricio Orozco-AlzateSang-Woon KimMarco LoogPublished in: S+SSPR (2014)
Keyphrases
- metric learning
- nearest neighbor
- distance function
- nearest neighbor classification
- distance metric
- distance metric learning
- k nearest neighbor
- semi supervised
- knn
- machine learning and pattern recognition
- pairwise
- dimensionality reduction
- person re identification
- learning tasks
- multi task
- data points
- high dimensional data
- high dimensional
- semi supervised clustering
- semi supervised learning
- mahalanobis metric
- training set
- prior knowledge
- background knowledge
- multiple features
- nearest neighbor classifier
- feature space
- machine learning
- subject to linear constraints
- semi definite programming
- labeled data
- distance measure
- learning algorithm