Generalization bounds for learning with linear, polygonal, quadratic and conic side knowledge

Published

Journal Article

© 2014, The Author(s). In this paper, we consider a supervised learning setting where side knowledge is provided about the labels of unlabeled examples. The side knowledge has the effect of reducing the hypothesis space, leading to tighter generalization bounds, and thus possibly better generalization. We consider several types of side knowledge, the first leading to linear and polygonal constraints on the hypothesis space, the second leading to quadratic constraints, and the last leading to conic constraints. We show how different types of domain knowledge can lead directly to these kinds of side knowledge. We prove bounds on complexity measures of the hypothesis space for quadratic and conic side knowledge, and show that these bounds are tight in a specific sense for the quadratic case.

Full Text

Duke Authors

Cited Authors

  • Tulabandhula, T; Rudin, C

Published Date

  • September 17, 2015

Published In

Volume / Issue

  • 100 / 2-3

Start / End Page

  • 183 - 216

Electronic International Standard Serial Number (EISSN)

  • 1573-0565

International Standard Serial Number (ISSN)

  • 0885-6125

Digital Object Identifier (DOI)

  • 10.1007/s10994-014-5478-4

Citation Source

  • Scopus