Pembangunan dimensi baru bagi model regresi eksponen (kreb): aplikasi dalam sains kesihatan
Loading...
Date
2020-09
Authors
Rohim, Rabi'atul 'Adawiyah Abdul
Journal Title
Journal ISSN
Volume Title
Publisher
Pusat Pengajian Sains Kesihatan, Universiti Sains Malaysia
Abstract
The exponential regression method is a non-linear regression method that is
commonly used in the field of biometrics. Generally, the exponential regression model
is often applied in research especially in estimating the growth rate of bacteria and
viruses. Similar to the linear regression method, the exponential regression model also
involves the relationship between two situations that can be explained in terms of
independent and dependent variables. Fundamentally, this non-linear relationship
(exponential relationship) can be embodied or transformed into a linear form (linear
equation). Through this transformation step, the efficiency of a model function can be
improved and enhanced. The main focus of this study is to establish a standard
procedure for integrated exponential function algorithm, the KREB Model, with
emphasis on the modelling techniques based on the scientific computational statistical
method and the advanced statistical method by utilising the SAS software. The KREB
Model is short for Integrated Exponential Regression Method. Essentially, this study
combines several advanced statistical methods that are integrated into a single SAS
programming language involving the exponential method, the bootstrap method as
well as fuzzy regression. An in-depth analysis of all three cases was carried out and
the results were compared thoroughly. The outcome from the KREB Model was
relatively compared with the existing methods. The findings from the comparison will
provide information regarding the effectiveness of the KREB Model. The KREB
Model has a smaller average interval than the existing model. The KREB Model
developed is better than the existing model and can make predictions more efficiently
and accurately. Future research proposals can be expanded by using exponentialregression methods by creating a KREB Model based on partial regregssion, spatial
regression and poison regression so that it can be used as a guide in future research.
In conclusion, the KREB Model that was developed has successfully optimised the
results of the study through the computational statistical calculation method.
Description
Keywords
biometrics