Multi-instance transfer metric learning by weighted distribution and consistent maximum likelihood estimation.
Siyu JiangYonghui XuHengjie SongQingyao WuMichael K. NgHuaqing MinShaojian QiuPublished in: Neurocomputing (2018)
Keyphrases
- metric learning
- multi instance
- maximum likelihood estimation
- probability distribution
- density function
- distance metric
- maximum likelihood
- em algorithm
- semi supervised
- parameter estimation
- learning tasks
- pairwise
- multi task
- semi supervised learning
- expectation maximization
- dimensionality reduction
- distance function
- transfer learning
- feature space
- probability density function
- euclidean distance
- pattern recognition
- least squares
- active learning