Nonlinear Laplacian spectral analysis: Capturing intermittent and low-frequency spatiotemporal patterns in high-dimensional data.
Dimitrios GiannakisAndrew J. MajdaPublished in: CoRR (2012)
Keyphrases
- spectral analysis
- low frequency
- high dimensional data
- graph laplacian
- underlying manifold
- spatiotemporal patterns
- high frequency
- frequency domain
- filter bank
- subband
- dimensionality reduction
- wavelet transform
- low dimensional
- high dimensional
- nearest neighbor
- subspace clustering
- discrete wavelet transform
- data sets
- wavelet coefficients
- clustering high dimensional data
- unique features
- data analysis
- autoregressive
- frequency band
- sparse representation
- data points
- multiresolution
- dimension reduction
- manifold learning
- random walk
- speech signal
- input data
- high resolution
- linear discriminant analysis
- edge detection
- neural network