Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal Bilangan: Jurnal Ilmiah Matematika, Kebumian dan Angkasa
Anwer Fawzi Ali
Unknown Affiliation
Astronomy Mathematics

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
ROBUST M-ESTIMATORS' Parameter Scale is Highly Efficient on Base Q Grades Zahraa Khaled Gaafar; Anwer Fawzi Ali
Bilangan : Jurnal Ilmiah Matematika, Kebumian dan Angkasa Vol. 2 No. 6 (2024): Bilangan : Jurnal Ilmiah Matematika, Kebumian dan Angkasa
Publisher : Asosiasi Riset Ilmu Matematika dan Sains Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62383/bilangan.v2i6.300

Abstract

Russiva and Crookes proposed the widely used, highly efficient, robust Q-score parameter scale in their 1993 work, and they approximated it using "fast" Huber's M-scores. Shown What proposed us M-scores are highly efficient and robust on arbitrary nom distributions, thanks to correct choice parameter approximations. We pay particular attention to the Cauchy divisions  and cases of Gaussian distribution . Important terms: Cauchy's law, parameter scale, robustness, and Gaussian distribution With the Q-estimate as a foundation, the M-estimates of the scale parameter are very efficient and robust. Rousseeuw and Croux (1993) proposed the scale parameter of Q-estimate, which is often used with fast Huber M-estimates, and we were able to get rather close to it. By meticulously choosing the approximation parameters, we proved that the proposed M-estimates are efficient and resilient for any data distribution. In order to measure the robustness and efficiency of scale M estimates , we calculated their asymptotic variances; breakdown points , and influence functions. The Cauchy and Gaussian distributions were our main focus. Notably, the proposed robust estimate is consistent the Maximum likelihood estimation for the Cauchy distribution. Last but not least, these robust and efficient scale estimates need three to four times less computing time than their comparable Q-estimates. The research provides robust and efficient solutions for estimating distribution parameters more quickly and with less complexity, making them suitable for applications requiring fast and accurate computations.  
Co-Authors Zahraa Khaled Gaafar