ONLINEKERNEL-BASEDCLASSIFICATION BY PROJECTIONS

Konstantinos Slavakis, Sergios Theodoridis

Abstract

Thegoal ofthis paper isthedevelopment ofanovel efficient online kernel-based algorithm forclassification. Thespirit ofthealgorithm stems fromtherecently introduced Adaptive Projected Subgradient Method. Thisisageneral convex analytic tool that employs projections ontoasequence ofconvex sets anditcanbeconsidered asa generalization ofthecelebrated APAalgorithm, widely usedinclassical adaptive filtering.
BibTeX
@inproceedings{slavakis2007onlinekernel,
  author = {Konstantinos Slavakis and Sergios Theodoridis},
  title = {{ONLINEKERNEL-BASEDCLASSIFICATION BY PROJECTIONS}},
  booktitle = {Tokyo},
  year = {2007},
}

Loading related content...

Figure