A kernel Principal Component Analysis (kPCA) digest with a new backward mapping (pre-image reconstruction) strategy.
Alberto García-GonzálezAntonio HuertaSergio ZlotnikPedro DíezPublished in: CoRR (2020)
Keyphrases
- image reconstruction
- kernel principal component analysis
- high dimensional feature space
- kernel pca
- principal components
- discriminant analysis
- feature extraction
- principal component analysis
- preprocessing
- kernel function
- classification method
- feature space
- kernel matrix
- kernel methods
- super resolution
- high dimensional
- maximum a posteriori
- face recognition
- feature vectors
- dimensionality reduction
- linear discriminant analysis
- data sets
- motion estimation
- knn
- image analysis
- random projections
- similarity measure
- subspace methods
- image processing