PERBANDINGAN METODE ROBUST ESTIMASI-LTS DENGAN ESTIMASI-S UNTUK DATA YANG MENGANDUNG PENCILAN (OUTLIER)

MUCH ARIF ABDULLAH, NIM. 14610042 (2018) PERBANDINGAN METODE ROBUST ESTIMASI-LTS DENGAN ESTIMASI-S UNTUK DATA YANG MENGANDUNG PENCILAN (OUTLIER). Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

[img] Text (PERBANDINGAN METODE ROBUST ESTIMASI-LTS DENGAN ESTIMASI-S UNTUK DATA YANG MENGANDUNG PENCILAN (OUTLIER))
14610042_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version
Restricted to Registered users only

Download (5MB)
[img] Text (PERBANDINGAN METODE ROBUST ESTIMASI-LTS DENGAN ESTIMASI-S UNTUK DATA YANG MENGANDUNG PENCILAN (OUTLIER))
14610042_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf - Published Version
Restricted to Registered users only

Download (4MB)

Abstract

Linear regression analysis is a technique in statistics that is used to form a model of relationship between the dependent variable with one or more independent variables. The method used to estimate the parameters in the regression analysis is Ordinary Least Square. Often the result obtained from this method is less precise to model the data contains outlier. One of the methods used when data contains outlier is by LTS-estimation robust regression method and Sestimation. LTS-estimation is robust estimate that uses the concept of trimming data distribution to minimize the least squares quantities to get the best regression model. Meanwhile, the S-estimation is an estimate that has a high breakdown point value of up to 50%, so the S-estimate can overcome half of the outliers and give a good effect on the resulting model. This study aims to compare the effectiveness of Ordinary Least Square Methods, LTS-estimation and Welsch-weighted S-estimation in forming regression models. Comparison of these methods is judged by the standard error and Adjusted R-Square. The case in this study was taken from West Java Provincial BPS on cassava production , harvested area , average rainfall , and productivity based on district and city in East Java Province 2014. Results of the study of 26 districts and cities, LTS-estimation resulted in a better model than the S-estimation and Ordinary Least Square Method. This is seen from the result of standard error of LTS-estimation, S-estimation, and OLS are 46.069, 546.2739, 19030 and Adjusted R-Square respectively 0.9999, 0.996, 0.9840.

Item Type: Thesis (Skripsi)
Additional Information: Dr. Epha Diana Supandi, M.Sc.,
Uncontrolled Keywords: LTS-Estimation, Ordinary Least Squares, Regression Analysis, SEstimation, Welsch.
Subjects: Matematika
Divisions: Fakultas Sains dan Teknologi > Matematika (S1)
Depositing User: H. Zaenal Arifin, S.Sos.I., S.IPI.
Date Deposited: 02 Jan 2019 11:23
Last Modified: 02 Jan 2019 11:23
URI: http://digilib.uin-suka.ac.id/id/eprint/32176

Share this knowledge with your friends :

Actions (login required)

View Item View Item
Chat Kak Imum