PENERAPAN METODE TRANSFORMASI LOGARITMA NATURAL DAN PARTIAL LEAST SQUARES UNTUK MEMPEROLEH MODEL BEBAS MULTIKOLINIER DAN OUTLIER

Ohyver, Margaretha (2013) PENERAPAN METODE TRANSFORMASI LOGARITMA NATURAL DAN PARTIAL LEAST SQUARES UNTUK MEMPEROLEH MODEL BEBAS MULTIKOLINIER DAN OUTLIER. Jurnal Mat Stat, 13 (01). pp. 42-51. ISSN 1412-1220

[img]
Preview
Text
05 Margaretha O- makalah logaritma pls - revisi-edit.pdf

Download (178kB) | Preview

Abstract

Multicollinear and outlier occur when making regression modeling. Multicollinear leads difficulty in separating the influence of each independent variable on the response variable. Outlier causes unmet assumption of normality in the regression. Both cases occur in the number of hotel visitors in Kendari. The purpose of this paper is to find a model that is free from multicollinear and outlier. Using the natural logarithm transformation and partial least squares, obtained model has the value of variance inflation factor less than ten and is able to overcome the outlier.

Item Type: Article
Subjects: MATHEMATICAL ANALYSIS
LINEAR PROGRAMMING
MATHEMATICS & STATISTICS
Depositing User: - Anindito
Date Deposited: 18 Apr 2013 04:42
Last Modified: 18 Apr 2013 04:42
URI: http://eprints.binus.ac.id/id/eprint/25629

Actions (login required)

View Item View Item