A Cubically Convergent Parallelizable Method for the Hermitian Eigenvalue Problem.
Hongyuan ZhaZhenyue ZhangPublished in: SIAM J. Matrix Anal. Appl. (1998)
Keyphrases
- high accuracy
- fully automatic
- objective function
- high precision
- computational cost
- cost function
- support vector machine
- mathematical model
- detection algorithm
- significant improvement
- classification accuracy
- classification method
- segmentation method
- input data
- clustering method
- computational complexity
- artificial neural networks
- preprocessing
- missing data
- error rate
- theoretical analysis
- feature set
- pairwise
- semi supervised
- dynamic programming
- prior knowledge