Recent developments in computer vision have shown that local features can provide efficient representations suitable for robust object recognition. Support Vector Machines have be...
Christian Wallraven, Barbara Caputo, Arnulf B. A. ...
In many applications it is desirable to learn from several kernels. "Multiple kernel learning" (MKL) allows the practitioner to optimize over linear combinations of kern...
Kernel rootkits can provide user level-malware programs with the additional capabilities of hiding their malicious activities by altering the legitimate kernel behavior of an oper...
Motivation: Remote homology detection between protein sequences is a central problem in computational biology. Supervised learning algorithms based on support vector machines are ...
Many unsupervised learning algorithms make use of kernels that rely on the Euclidean distance between two samples. However, the Euclidean distance is optimal for Gaussian distribut...
Karim T. Abou-Moustafa, Mohak Shah, Fernando De la...