Abstract: Kernel classifiers based on Support Vector Machines (SVM) have achieved state-ofthe-art results in several visual classification tasks, however, recent publications and d...
Guo ShengYang, Min Tan, Si-Yao Fu, Zeng-Guang Hou,...
Recently, Balcan and Blum [1] suggested a theory of learning based on general similarity functions, instead of positive semi-definite kernels. We study the gap between the learnin...
Kernel functions have become an extremely popular tool in machine learning, with an attractive theory as well. This theory views a kernel as implicitly mapping data points into a ...
SockMi is a solution for the migration of TCP/IP connections between Linux systems. Only the migrating peer of the connection needs to reside on a Linux system. The migration is c...
Massimo Bernaschi, Francesco Casadei, Paolo Tassot...
Gaussian kernels with flexible variances provide a rich family of Mercer kernels for learning algorithms. We show that the union of the unit balls of reproducing kernel Hilbert s...