Rule extraction from a trained neural network for image keywords extraction

H. Nishiyama, H. Kawasaki, M. Fukumi, N. Akamatsu, Y. Mitsukura

Research output: Contribution to conferencePaperpeer-review

Abstract

This paper presents a rule extraction method from a trained neural network (NN), which is used for keywords extraction from Images. In our approach, first, a bit map image in the RGB color space is transformed into that in the L*a*b* color space. Next, it clusters image pixels using the fuzzy c-means method and domains are extracted through a labeling process. Features, such as area of obtained domains, color information, and coordinates of the center of gravity, are then calculated, which are used as input attributes to NN. NN is then trained using such features. After NN learning, rule extraction is carried out using binarized output values in the hidden layer for each keyword. The rules extracted in this paper are If-then rules, which include logical functions. The methods of generating keywords using NN and the rules are presented and their comparative experiments are performed. Finally the validity of these methods was verified by means of computer simulations.

Original languageEnglish
Pages325-329
Number of pages5
DOIs
Publication statusPublished - 2004 Jan 1
Externally publishedYes
EventNAFIPS 2004 - Annual Meeting of the North American Fuzzy Information Processing Society: Fuzzy Sets in the Heart of the Canadian Rockies - Banff, Alta, Canada
Duration: 2004 Jun 272004 Jun 30

Other

OtherNAFIPS 2004 - Annual Meeting of the North American Fuzzy Information Processing Society: Fuzzy Sets in the Heart of the Canadian Rockies
CountryCanada
CityBanff, Alta
Period04/6/2704/6/30

ASJC Scopus subject areas

  • Computer Science(all)
  • Mathematics(all)

Fingerprint Dive into the research topics of 'Rule extraction from a trained neural network for image keywords extraction'. Together they form a unique fingerprint.

Cite this