Background: Protein remote homology detection is a central problem in computational biology. Most recent methods train support vector machines to discriminate between related and ...
The kernel-parameter is one of the few tunable parameters in Support Vector machines, controlling the complexity of the resulting hypothesis. Its choice amounts to model selection...
Nello Cristianini, Colin Campbell, John Shawe-Tayl...
Extending recognition to uncontrolled situations is a key challenge for practical face recognition systems. Finding efficient and discriminative facial appearance descriptors is c...
It is difficult to adapt discriminative classifiers, particularly kernel based ones such as support vector machines (SVMs), to handle mismatches between the training and test da...
Motivated by the setting of reproducing kernel Hilbert space (RKHS) and its extensions considered in machine learning, we propose an RKHS framework for image and video colorizatio...