Likelihood-based non-uniform allocation of Gaussian kernels in scalar dimension for HMM compression.
Xiao-Bing LiDouglas D. O'ShaughnessyPublished in: ICME (2008)
Keyphrases
- gaussian kernels
- hidden markov models
- kernel function
- learning rate
- covering numbers
- gaussian kernel
- reproducing kernel hilbert space
- maximum likelihood
- feature space
- high dimensional feature space
- multiple kernel learning
- convergence rate
- kernel methods
- least squares
- input space
- binary classification
- scale space
- special case
- support vector