dc.description.abstract |
The purpose of this study was to predict academic achievement among engineering students
and to pinpoint the elements or characteristics that influence this performance. To predict
academic achievement, the study employed several machine learning models such as Linear
Regression, Random Forest, Xgbooster, Artificial neural network and an ensemble of 3, and
then compared their performance to find the best model. The study also looked at the
important variables that affect academic accomplishment, such as demography,
socioeconomic position, high school academic performance, and other pertinent variables.
The study's conclusions could enhance academic support, counselling for engineering
students, and instructional strategies.
An ensemble model surpassed any individual machine learning model, according to the
study, which assessed the accuracy and precision of several machine learning models.
Furthermore, the study found that past academic success in particular disciplines, such as
Biology, English Language, Critical Reading, Citizen Competencies, and Mathematics,
significantly influenced the academic performance of engineering students. However, while
high schools and institutions had a positive or negative impact, the socioeconomic
background of the students had no discernible impact on their academic achievement. While
having no influence on female students, the demographic factor of gender had a beneficial
effect on the academic performance of male students.
The lack of information on the students' academic achievement across various subject areas
throughout their time at university also presented problems for the study. Making more
precise prediction in the future will depend on gathering more detailed data on students'
academic performance across various university-level courses. This knowledge might aid in
developing better strategies for increasing academic outcomes in certain areas and help us
better understand how certain disciplines impact academic achievement |
en_US |