Some convergence properties of iterated reweighted least squares in the location model
作者:
Birch B. Jeffrey,
期刊:
Communications in Statistics - Simulation and Computation
(Taylor Available online 1980)
卷期:
Volume 9,
issue 4
页码: 359-369
ISSN:0361-0918
年代: 1980
DOI:10.1080/03610918008812162
出版商: Marcel Dekker, Inc.
关键词: m-estimators;redescending ip-functions;rates of convergence
数据来源: Taylor
摘要:
The concept of the M-estimator applied to the location model has been studied extensively under a variety of departures from the classical location model. One commonly used computational technique for obtaining the M-estimate is the iterated reweighted least squares algorithm. The convergence properties of this method will be discussed and compared to the well known Newton's method. Some of the properties of both methods will be illustrated in a simple example.
点击下载:
PDF (238KB)
返 回