Transductive Rademacher Complexity and Its Applications

Abstract

We present data-dependent error bounds for transductive learning based on transductive Rademacher complexity. For specific algorithms we provide bounds on their Rademacher complexity based on their "unlabeled-labeled" decomposition. This decomposition technique applies to many current and practical graph-based algorithms. Finally, we present a new PAC-Bayesian bound for mixtures of transductive algorithms based on our Rademacher bounds.

Cite

Text

El-Yaniv and Pechyony. "Transductive Rademacher Complexity and Its Applications." Journal of Artificial Intelligence Research, 2009. doi:10.1613/JAIR.2587

Markdown

[El-Yaniv and Pechyony. "Transductive Rademacher Complexity and Its Applications." Journal of Artificial Intelligence Research, 2009.](https://mlanthology.org/jair/2009/elyaniv2009jair-transductive/) doi:10.1613/JAIR.2587

BibTeX

@article{elyaniv2009jair-transductive,
  title     = {{Transductive Rademacher Complexity and Its Applications}},
  author    = {El-Yaniv, Ran and Pechyony, Dmitry},
  journal   = {Journal of Artificial Intelligence Research},
  year      = {2009},
  pages     = {193-234},
  doi       = {10.1613/JAIR.2587},
  volume    = {35},
  url       = {https://mlanthology.org/jair/2009/elyaniv2009jair-transductive/}
}