Probabilistic learnability of context-free grammars with basic distributional properties from positive examples

Chihiro Shibata, Ryo Yoshinaka

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

In recent years different interesting subclasses of cfls have been found to be learnable by techniques generically called distributional learning. The theoretical study on the exact learning of cfls by those techniques under different learning schemes is now quite mature. On the other hand, positive results on the pac learnability of cfls are rather limited and quite weak. This paper shows that several subclasses of context-free languages that are known to be exactly learnable with positive data and membership queries by distributional learning techniques are pac learnable from positive data under some assumptions on the string distribution.

Original languageEnglish
Pages (from-to)46-72
Number of pages27
JournalTheoretical Computer Science
Volume620
DOIs
Publication statusPublished - 2016 Mar 21
Externally publishedYes

Keywords

  • Distributional learning
  • Grammatical inference
  • PAC learning

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Probabilistic learnability of context-free grammars with basic distributional properties from positive examples'. Together they form a unique fingerprint.

  • Cite this