Multi-Output Learning via Spectral Filtering

Abstract

In this paper we study a class of regularized kernel methods for multi-output learning which are based on filtering the spectrum of the kernel matrix. The considered methods include Tikhonov regularization as a special case, as well as interesting alternatives such as vector-valued extensions of L2 boosting and other iterative schemes. Computational properties are discussed for various examples of kernels for vector-valued functions and the benefits of iterative techniques are illustrated. Generalizing previous results for the scalar case, we show a finite sample bound for the excess risk of the obtained estimator, which allows to prove consistency both for regression and multi-category classification. Finally, we present some promising results of the proposed algorithms on artificial and real data.

Cite

Text

Baldassarre et al. "Multi-Output Learning via Spectral Filtering." Machine Learning, 2012. doi:10.1007/S10994-012-5282-Y

Markdown

[Baldassarre et al. "Multi-Output Learning via Spectral Filtering." Machine Learning, 2012.](https://mlanthology.org/mlj/2012/baldassarre2012mlj-multioutput/) doi:10.1007/S10994-012-5282-Y

BibTeX

@article{baldassarre2012mlj-multioutput,
  title     = {{Multi-Output Learning via Spectral Filtering}},
  author    = {Baldassarre, Luca and Rosasco, Lorenzo and Barla, Annalisa and Verri, Alessandro},
  journal   = {Machine Learning},
  year      = {2012},
  pages     = {259-301},
  doi       = {10.1007/S10994-012-5282-Y},
  volume    = {87},
  url       = {https://mlanthology.org/mlj/2012/baldassarre2012mlj-multioutput/}
}