Publication: Suatu kaedah alternatif bagi pemodelan linear dalam biostatistik
Loading...
Date
2018-09
Authors
Ibrahim, Mohamad Shafiq Mohd
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
model. Current technological advancement and increase in development of
the new or modified methodology building leads to the development of an alternative
method for multiple linear regression model calculation. In this study, multiple linear
regression model will be calculated by using SAS programming language based on
computational statistics which will consider combination of robust regression,
bootstrap, weighted data, Bayesian and fuzzy regression method. Methodology
building is based on the SAS algorithm (SAS 9.4 software) which is a robust
computational statistics that consists of the combination of robust regression,
bootstrap, weighted data, bayesian and fuzzy regression methods. Three different SAS
algorithms that is; (i) Bootstrap Multiple Linear Regression (BMLR), (ii) Bootstrap
Weighted Bayesian Multiple Linear Regression (BWBMLR) and (iii) Fuzzy Bootstrap
Weighted Multiple Linear Regression (FBWMLR) will be compared separately
according to their average width of prediction. To illustrate the potential of built-in
algorithm, three cases of study will be used which are modeling on systolic blood
pressure level, modeling on tumour size and modeling on Early Childhood Caries
(ECC). The average width of prediction interval results for all cases of the models has
been computed and compared. The smallest width will indicate the best fitting model.
The result shows that the former MLR model has an average width of 11.2948, 7.4816
and 29.0407; and BMLR model has an average width of 2.5785, 3.7098 and 19.6589.
Description
Keywords
Linear models