An empirical approach for probing the definiteness of kernels.
Martin ZaeffererThomas Bartz-BeielsteinGünter RudolphPublished in: Soft Comput. (2019)
Keyphrases
- positive definite
- kernel methods
- kernel function
- support vector
- feature space
- linear combination
- multiple kernel
- covariance matrix
- search engine
- data structure
- geometric structure
- reproducing kernel hilbert space
- svm classification
- gaussian kernels
- real world
- artificial neural networks
- loss function
- objective function
- image processing
- gaussian processes
- multiple kernel learning
- knowledge base