An Inequality for Nearly Log-Concave Distributions with Applications to Learning

Abstract

We prove that given a nearly log-concave density, in any partition of the space to two well separated sets, the measure of the points that do not belong to these sets is large. We apply this isoperimetric inequality to derive lower bounds on the generalization error in learning. We also show that when the data are sampled from a nearly log-concave distribution, the margin cannot be large in a strong probabilistic sense. We further consider regression problems and show that if the inputs and outputs are sampled from a nearly log-concave distribution, the measure of points for which the prediction is wrong by more than ε _0 and less than ε _1 is (roughly) linear in ε _1– ε _0.

Cite

Text

Caramanis and Mannor. "An Inequality for Nearly Log-Concave Distributions with Applications to Learning." Annual Conference on Computational Learning Theory, 2004. doi:10.1007/978-3-540-27819-1_37

Markdown

[Caramanis and Mannor. "An Inequality for Nearly Log-Concave Distributions with Applications to Learning." Annual Conference on Computational Learning Theory, 2004.](https://mlanthology.org/colt/2004/caramanis2004colt-inequality/) doi:10.1007/978-3-540-27819-1_37

BibTeX

@inproceedings{caramanis2004colt-inequality,
  title     = {{An Inequality for Nearly Log-Concave Distributions with Applications to Learning}},
  author    = {Caramanis, Constantine and Mannor, Shie},
  booktitle = {Annual Conference on Computational Learning Theory},
  year      = {2004},
  pages     = {534-548},
  doi       = {10.1007/978-3-540-27819-1_37},
  url       = {https://mlanthology.org/colt/2004/caramanis2004colt-inequality/}
}