Likelihood-ratio empirical kernels for i-vector based PLDA-SVM scoring.
Man-Wai MakWei RaoPublished in: ICASSP (2013)
Keyphrases
- likelihood ratio
- support vector
- kernel function
- hypothesis testing
- svm classification
- feature vectors
- kernel methods
- support vector machine svm
- polynomial kernels
- support vector machine
- multiple kernel learning
- feature space
- linear discriminant analysis
- hypothesis test
- multiple kernel
- rbf kernel
- dot product
- knn
- gaussian kernel
- string kernels
- statistical learning theory
- support vectors
- kernel machines
- kernel parameters
- feature selection
- linear svm
- standard svm
- hyperplane
- svm classifier
- training data
- match scores
- multi class
- machine learning
- likelihood ratio test
- positive definite
- kernel svms
- histogram intersection kernel
- tree kernels
- gaussian kernels
- reproducing kernel hilbert space
- learning machines
- binary classification
- confidence intervals
- generalization ability
- test statistic
- radial basis function
- loss function
- upper bound
- data mining