RdimKD: Generic Distillation Paradigm by Dimensionality Reduction.
Yi GuoYiqian HeXiaoyang LiHaotong QinVan Tung PhamYang ZhangShouda LiuPublished in: CoRR (2023)
Keyphrases
- dimensionality reduction
- principal component analysis
- feature extraction
- data representation
- pattern recognition
- high dimensional
- input space
- high dimensional data
- structure preserving
- dimension reduction
- high dimensionality
- manifold learning
- low dimensional
- domain specific
- high level
- principal components
- feature selection
- preprocessing step
- pattern recognition and machine learning
- euclidean distance
- data points
- data sets
- linear dimensionality reduction
- singular value decomposition
- dimensionality reduction methods
- kernel pca
- random projections
- linear discriminant analysis
- database
- real time