Hardware architecture for positive definite matrix inversion based on LDL decomposition and back-substitution.
Carl IngemarssonOscar GustafssonPublished in: ACSSC (2016)
Keyphrases
- positive definite
- matrix inversion
- hardware architecture
- kernel matrix
- kernel function
- kernel methods
- processing elements
- hardware implementation
- low rank
- feature space
- covariance matrix
- least squares
- metric learning
- model selection
- input space
- associative memory
- training samples
- field programmable gate array
- semi parametric
- neural network
- convex optimization
- ls svm
- support vectors
- principal component analysis
- pattern recognition
- support vector
- learning algorithm
- learning tasks
- missing data
- machine learning