Generalization bounds for learning weighted automata

Borja Balle, Mehryar Mohri

Research output: Contribution to journalArticle

Abstract

This paper studies the problem of learning weighted automata from a finite sample of strings with real-valued labels. We consider several hypothesis classes of weighted automata defined in terms of three different measures: the norm of an automaton's weights, the norm of the function computed by an automaton, and the norm of the corresponding Hankel matrix. We present new data-dependent generalization guarantees for learning weighted automata expressed in terms of the Rademacher complexity of these classes. We further present upper bounds on these Rademacher complexities, which reveal key new data-dependent terms related to the complexity of learning weighted automata.

Original languageEnglish (US)
JournalTheoretical Computer Science
DOIs
StateAccepted/In press - Jan 1 2017

Fingerprint

Weighted Automata
Learning Automata
Labels
Dependent Data
Norm
Automata
Hankel Matrix
Strings
Upper bound
Generalization
Term
Class

Keywords

  • Generalization bounds
  • Learning theory
  • Rademacher complexity
  • Weighted automata

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Generalization bounds for learning weighted automata. / Balle, Borja; Mohri, Mehryar.

In: Theoretical Computer Science, 01.01.2017.

Research output: Contribution to journalArticle

@article{b82e7618ad0041d8a0a095e217a2fdd2,
title = "Generalization bounds for learning weighted automata",
abstract = "This paper studies the problem of learning weighted automata from a finite sample of strings with real-valued labels. We consider several hypothesis classes of weighted automata defined in terms of three different measures: the norm of an automaton's weights, the norm of the function computed by an automaton, and the norm of the corresponding Hankel matrix. We present new data-dependent generalization guarantees for learning weighted automata expressed in terms of the Rademacher complexity of these classes. We further present upper bounds on these Rademacher complexities, which reveal key new data-dependent terms related to the complexity of learning weighted automata.",
keywords = "Generalization bounds, Learning theory, Rademacher complexity, Weighted automata",
author = "Borja Balle and Mehryar Mohri",
year = "2017",
month = "1",
day = "1",
doi = "10.1016/j.tcs.2017.11.023",
language = "English (US)",
journal = "Theoretical Computer Science",
issn = "0304-3975",
publisher = "Elsevier",

}

TY - JOUR

T1 - Generalization bounds for learning weighted automata

AU - Balle, Borja

AU - Mohri, Mehryar

PY - 2017/1/1

Y1 - 2017/1/1

N2 - This paper studies the problem of learning weighted automata from a finite sample of strings with real-valued labels. We consider several hypothesis classes of weighted automata defined in terms of three different measures: the norm of an automaton's weights, the norm of the function computed by an automaton, and the norm of the corresponding Hankel matrix. We present new data-dependent generalization guarantees for learning weighted automata expressed in terms of the Rademacher complexity of these classes. We further present upper bounds on these Rademacher complexities, which reveal key new data-dependent terms related to the complexity of learning weighted automata.

AB - This paper studies the problem of learning weighted automata from a finite sample of strings with real-valued labels. We consider several hypothesis classes of weighted automata defined in terms of three different measures: the norm of an automaton's weights, the norm of the function computed by an automaton, and the norm of the corresponding Hankel matrix. We present new data-dependent generalization guarantees for learning weighted automata expressed in terms of the Rademacher complexity of these classes. We further present upper bounds on these Rademacher complexities, which reveal key new data-dependent terms related to the complexity of learning weighted automata.

KW - Generalization bounds

KW - Learning theory

KW - Rademacher complexity

KW - Weighted automata

UR - http://www.scopus.com/inward/record.url?scp=85036514806&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85036514806&partnerID=8YFLogxK

U2 - 10.1016/j.tcs.2017.11.023

DO - 10.1016/j.tcs.2017.11.023

M3 - Article

JO - Theoretical Computer Science

JF - Theoretical Computer Science

SN - 0304-3975

ER -