Nonlinear Laplacian spectral analysis: capturing intermittent and low-frequency spatiotemporal patterns in high-dimensional data.
Dimitrios GiannakisAndrew J. MajdaPublished in: Stat. Anal. Data Min. (2013)
Keyphrases
- spectral analysis
- low frequency
- high dimensional data
- graph laplacian
- underlying manifold
- spatiotemporal patterns
- high frequency
- subband
- frequency domain
- filter bank
- high dimensional
- nearest neighbor
- wavelet transform
- low dimensional
- data sets
- dimensionality reduction
- data analysis
- discrete wavelet transform
- unique features
- subspace clustering
- wavelet coefficients
- clustering high dimensional data
- frequency band
- data points
- linear discriminant analysis
- manifold learning
- feature vectors
- image compression
- dimension reduction
- autoregressive
- speech signal
- sparse representation
- random fields
- bit rate
- signal processing
- edge detection
- high resolution
- multiresolution
- multiscale
- feature selection