2024-03-28T09:27:52Z
https://meral.edu.mm/oai
oai:meral.edu.mm:recid/1383
2024-02-27T08:24:08Z
1582963436320
1582963436320:1582965639643
user-yueco
An Application of Ordinary Least Squares and Maximum Likelihood Type Estimation in Roust Diagnostic Regression Analysis (Maw Maw Khin, 2011)
MAW MAW KHIN
This study shows that the OLS method is quite sensitive to outlier whereas maximum likelihood type estimation {M-estimation) methods resist outliers. The iteratived reweighted least squares (IRLS) method based on the Huber and the Bisquare 'I' -functions clearly detect outliers that are given to less weight. The findings show that
maximum likelihood type estimation based on the mean squares error (MSE) criterion can provide predicted values very close to actual values.
2011-01-01
http://hdl.handle.net/20.500.12678/0000001383
https://meral.edu.mm/records/1383