Parameter Estimation for Pareto Distribution and Type-II Fuzzy Logic


ERBAY DALKILIÇ T. , ŞANLI KULA K.

GAZI UNIVERSITY JOURNAL OF SCIENCE, vol.30, no.1, pp.251-258, 2017 (Journal Indexed in ESCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 30 Issue: 1
  • Publication Date: 2017
  • Title of Journal : GAZI UNIVERSITY JOURNAL OF SCIENCE
  • Page Numbers: pp.251-258

Abstract

While parameter estimation is done by the classical methods, there are a number of assumptions need to be satisfied, in the linear regression analysis. Key assumptions of linear regression are; no auto correlation, no or little multicollinearity, homoscedasticity and the errors have normal distribution. In this work, the case that independent variable has Pareto distribution to be discussed and an algorithm using adaptive networks suggested to parameter estimation where the.. which is one of the parameters of the fuzzy membership functions is fuzzy. Also the parameter of fuzzy membership function is fuzzy the estimation process is based on type-II fuzzy logic.