TY - GEN

T1 - On the complexity of samples for learning

AU - Ratsaby, Joel

PY - 2004

Y1 - 2004

N2 - In machine-learning, maximizing the sample margin can reduce the learning generalization-error. Thus samples on which the target function has a large margin (γ) convey more information so we expect fewer such samples. In this paper, we estimate the complexity of a class of sets of large-margin samples for a general learning problem over a finite domain. We obtain an explicit dependence of this complexity on γ and the sample size.

AB - In machine-learning, maximizing the sample margin can reduce the learning generalization-error. Thus samples on which the target function has a large margin (γ) convey more information so we expect fewer such samples. In this paper, we estimate the complexity of a class of sets of large-margin samples for a general learning problem over a finite domain. We obtain an explicit dependence of this complexity on γ and the sample size.

UR - http://www.scopus.com/inward/record.url?scp=35048871696&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-27798-9_23

DO - 10.1007/978-3-540-27798-9_23

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:35048871696

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 198

EP - 209

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

A2 - Chwa, Kyung-Yong

A2 - Ian Munro, J.

ER -