On approximately identifying concept classes in the limit

Satoshi Kobayashi, Takashi Yokomori

Research output: Chapter in Book/Report/Conference proceedingConference contribution

23 Citations (Scopus)

Abstract

In this paper, we introduce various kinds of approximations of a concept and propose a framework of approximate learning in case that a target concept could be outside the hypothesis space. We present some char­acterization theorems for approximately identifiability. In particular, we show a remarkable result that the upper-best approximate identifiability from com­plete data is collapsed into the upper-best approximate identifiability from positive data. Further, some other characterizations for approximate identifi­ability from positive data are presented, where we establish a relationship be­tween approximate identifiability and some important notions in quasi-order theory and topology theory. The results obtained in this paper are essentially related to the closure property of concept classes under infinite intersections (or infinite unions). We also show that there exist some interesting example concept classes with such properties (including specialized EFS’s) by which an upper-best approximation of any concept can be identifiable in the limit from positive data.

Original languageEnglish
Title of host publicationAlgorithmic Learning Theory - 6th International Workshop, ALT 1995, Proceedings
EditorsKlaus P. Jantke, Takeshi Shinohara, Thomas Zeugmann
PublisherSpringer Verlag
Pages298-312
Number of pages15
ISBN (Print)3540604545, 9783540604549
DOIs
Publication statusPublished - 1995
Event6th International Workshop on Algorithmic Learning Theory, ALT 1995 - Fukuoka, Japan
Duration: 1995 Oct 181995 Oct 20

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume997
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other6th International Workshop on Algorithmic Learning Theory, ALT 1995
CountryJapan
CityFukuoka
Period95/10/1895/10/20

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'On approximately identifying concept classes in the limit'. Together they form a unique fingerprint.

  • Cite this

    Kobayashi, S., & Yokomori, T. (1995). On approximately identifying concept classes in the limit. In K. P. Jantke, T. Shinohara, & T. Zeugmann (Eds.), Algorithmic Learning Theory - 6th International Workshop, ALT 1995, Proceedings (pp. 298-312). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 997). Springer Verlag. https://doi.org/10.1007/3-540-60454-5_47