ASKIT: An Efficient, Parallel Library for High-Dimensional Kernel Summations.
William B. MarchBo XiaoChenhan D. YuGeorge BirosPublished in: SIAM J. Sci. Comput. (2016)
Keyphrases
- high dimensional
- feature space
- kernel space
- kernel function
- dimensionality reduction
- low dimensional
- input space
- shared memory
- kernel methods
- noisy data
- additive models
- high dimensional problems
- parallel implementation
- feature vectors
- similarity function
- parallel processing
- high dimensionality
- variable selection
- similarity search
- high dimensional data
- sparse data
- computationally efficient
- parallel programming
- positive definite
- high dimensional feature space
- kernel principal component analysis
- neural network