Fuzzy Least Squares Regression Model Based of Weighted Distance between Fuzzy Numbers

Nasibov E.

AUTOMATIC CONTROL AND COMPUTER SCIENCES, cilt.41, sa.1, ss.10-17, 2007 (ESCI) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 41 Sayı: 1
  • Basım Tarihi: 2007
  • Doi Numarası: 10.3103/s0146411607010026
  • Dergi Adı: AUTOMATIC CONTROL AND COMPUTER SCIENCES
  • Derginin Tarandığı İndeksler: Emerging Sources Citation Index (ESCI), Scopus
  • Sayfa Sayıları: ss.10-17
  • Anahtar Kelimeler: WABL, fuzzy number, fuzzy distance, fuzzy regression
  • Dokuz Eylül Üniversitesi Adresli: Evet

Özet

A new distance ND1 between fuzzy numbers based on an averaged representative of a fuzzy number (Weighting Average Based on Levels, WABL) is proposed. Based on this distance a new least squares regression model is proposed. The proposed model is studied for a broad class of fuzzy numbers and class of functions the membership of which is formed on the basis of the template mu([.])(x) = max(0.1 - vertical bar x vertical bar(s)).