Generalization bounds for learning with linear, polygonal, quadratic and conic side knowledge
Name
10994_2014_5478_ReferencePDF.pdf
Size
521.87 KB
Format
Adobe PDF
Checksum (MD5)
3a83a8dea10e8a392392008de08b5d20
Author(s)
Tulabandhula, Theja
Rudin, Cynthia
Date Issued
December 2014
Journal
Machine Learning
Publisher
Springer Science+Business Media
Citation
Tulabandhula, Theja, and Cynthia Rudin. "Generalization bounds for learning with linear, polygonal, quadratic and conic side knowledge." Machine Learning 100:2-3 (2015), pp.183-216.
Version
Author's final manuscript
Abstract
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.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Sloan School of Management
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1007/s10994-014-5478-4
