2023-01-29T05:24:13Zhttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_oaipmhoai:ipsj.ixsq.nii.ac.jp:000141282022-10-21T05:24:51Z00581:00768:00772
Parameter Estimation in the Extreme - Value Distributions Using the Continuation MethodParameter Estimation in the Extreme - Value Distributions Using the Continuation Methodeng論文http://id.nii.ac.jp/1001/00014128/Journal Articlehttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_action_common_download&item_id=14128&item_no=1&attribute_id=1&file_no=1Copyright (c) 1994 by the Information Processing Society of Japan情報数学Info. Math. Res. Lab. Takaoka Electric ; Fac. Info. Sci. Hiroshima City UniversityHideo, HiroseAn efficient and stable maximum likelihood parameter estimation scheme is introduced for the three kinds of extreme-value distribution (Weibull Gumbel and Frechet) using the generalized extreme-value distribution and the continuation method. As the proposed algorithm can almost always obtain the existing local maximum likelihood estimates automatically it is of considerable practical value. This paper focuses on the Weibull distribution parameter estimation and shows that it is better to use the generalized extreme-value distribution than the Weibull distribution itself and that the continuation method is more efficient than the grid search method in searching for parameters globally. The paper also shows that when there are no finite local maximum likelihood estimates in the Weibull distribution it is probable that there are finite local maxlmum likelihood estimates in the Frechet distribution and vice versa. Only complete data sets are considered in this paper but the algorithm can easily be applied to censored data.An efficient and stable maximum likelihood parameter estimation scheme is introduced for the three kinds of extreme-value distribution (Weibull, Gumbel, and Frechet) using the generalized extreme-value distribution and the continuation method. As the proposed algorithm can almost always obtain the existing local maximum likelihood estimates automatically, it is of considerable practical value. This paper focuses on the Weibull distribution parameter estimation and shows that it is better to use the generalized extreme-value distribution than the Weibull distribution itself, and that the continuation method is more efficient than the grid search method in searching for parameters globally. The paper also shows that when there are no finite local maximum likelihood estimates in the Weibull distribution, it is probable that there are finite local maxlmum likelihood estimates in the Frechet distribution, and vice versa. Only complete data sets are considered in this paper, but the algorithm can easily be applied to censored data.AN00116647情報処理学会論文誌359167416811994-09-151882-77642009-06-29