ISSN: 1813-0410

Author : D. Mohamed, Karrar

Linear Regression Model Related to Gumbel Distribution

Ahmed Al-Adilee; Karrar D. Mohamed

journal of kerbala university, Volume 10, Issue 2, Pages 70-78

Linear regression model is proposed to construct a proper estimation of parameters of Gumbel distribution as influences that effect on manufactured items, for example, bulbs, electrical machines, and so on. Such type of distributions and its properties that relevant to Gamma-distribution has been shown in this study. Some contributions of other extreme value distributions types such as Weibull distribution, and Fre ́chet distribution were introduced in general forms as models used for lifetime data. Maximum likelihood estimation (M.L.E) method was suggested to estimate the values of parameters because of its effective ability to obtain the best approximation for fitting data compared with other estimation methods like least square method, moments method, and other estimation methods. Another important fact of choosing this method is that this method regarded as a best method for small sample sizes.