Anomaly Detection with Score Functions Based on the Reconstruction Error of the Kernel PCA.
Laetitia ChapelChloé FriguetPublished in: ECML/PKDD (1) (2014)
Keyphrases
- anomaly detection
- reconstruction error
- kernel pca
- singular value decomposition
- kernel methods
- principal component analysis
- sparse representation
- sparse coding
- feature space
- dimensionality reduction
- subband
- spectral clustering
- objective function
- filter bank
- kernel matrix
- feature extraction
- face recognition
- preprocessing
- neural network
- kernel function
- unsupervised learning