Gnophanxay Chaylasy. A combination of Bayesian belief network and genetic algorithm for students' achievements of Faculty of Science National University of Laos. Master's Degree(Information Technology). King Mongkut's University of Technology North Bangkok. Central Library. : King Mongkut's University of Technology North Bangkok, 2011.
A combination of Bayesian belief network and genetic algorithm for students' achievements of Faculty of Science National University of Laos
Abstract:
Students achievement is the main target of education. According to statistics of
academic division of Faculty of Science (FOS), National University of Laos (NUOL)
during 2004-2008, 54% of students failed when they took the pre-course. This issue
impacts to academic division, so a need to find factors that influenced this outcome is
necessary. Once the relationship is discovered among subjects of the pre-course, those
subjects should be investigated to understand the students achievement or lack of
achievement. This thesis applies a combination of Bayesian Belief Network (BBN)
and Genetic Algorithm (GA). The BBN is a Directed Acyclic Graphical (DAG)
model that represents a set of random variables and their conditional dependencies via
a DAG, which makes it perfect for this research. The GA is also a famous of
evolutionary algorithm, and can be used as a search algorithm for BBN structure
learning. In this thesis, the researcher modified the mutation operator of GA to
improve the performance of BBN. The experimental data sets obtained from the
student score database of FOS pre-course during academic year 2004 to 2008. From
the result of the experiment, it is found that when using 2 point mutation the best
result was achieved compared to traditional methods.