A Class of New Kernels Based on High-Scored Pairs of k-Peptides for SVMs and Its Application for Prediction of Protein Subcellular Localization.
Zhengdeng LeiYang DaiPublished in: Trans. Comp. Sys. Biology (2005)
Keyphrases
- support vector
- kernel function
- binding peptides
- linear svm
- multi class classification
- prediction accuracy
- multi class
- neural networks and support vector machines
- multiclass classification
- feature selection
- kernel methods
- prediction model
- multiple kernel learning
- machine learning
- svm classifier
- binary classification
- protein interaction
- linear combination
- training data
- polynomial kernels
- gaussian kernels
- training set
- kernel machines
- support vector machine
- support vector machine svm
- learning machines
- support vectors
- gaussian processes
- loss function