A Periodic qd-Type Reduction for Computing Eigenvalues of Structured Matrix Products to High Relative Accuracy.
Rong HuangPublished in: J. Sci. Comput. (2018)
Keyphrases
- wide range
- computational cost
- covariance matrix
- singular value decomposition
- error rate
- high accuracy
- prediction accuracy
- eigenvalues and eigenvectors
- principal component analysis
- data sets
- computational efficiency
- high dimensional
- error reduction
- least squares
- high precision
- classification accuracy
- highly accurate
- matrix representation