A system to predict pharmaceutical student performance using machine learning

Nurul Nazihah, Sukor (2019) A system to predict pharmaceutical student performance using machine learning. Faculty of Computer System & Software Engineering, Universiti Malaysia Pahang.

17.A system to predict pharmaceutical student performance using machine learning.pdf - Accepted Version

Download (975kB) | Preview


Recently, pharmacy course is the important course as it involves human lives. Moreover, predicting of student performance is most of higher learning institutions in Malaysia. The main objective of this paper is to provide an overview on the Machine Learning technique that has been used to propose to improve student achievement and to predict the pharmacy student will be quit or graduate of a University. There are three reasons why this is happening relate in the problem statement. Firstly is the lack of pharmacy student in the sector in the pharmaceutical industry. Second is increase the number of student from local IPT to offer a pharmacy programed and lastly is the result of pharmacist shortage that student needed. As well know, if they take the pharmaceutical subject whether they have basic or not in the pharmaceutical, the result will display many pharmacy students is failed. This is because, they still don‟t know that they still suitable with the course or not. Here, the research is very important and suitable to needs to implement a system to predict pharmaceutical student performance using machine learning. This paper also focuses on how prediction uses Machine Learning classifier and finalized model to make a prediction on new data before the student registering the pharmaceutical course from WEKA to Java code. It could bring the benefit and impact to the student. This study proposes a Machine Learning technique to predict either the pharmacy student will be quit or graduate. The classifiers are multilayer perceptron (MLP) to predict pharmaceutical student performance. Among the MLP classifier, the outstanding outcome acquire is the MLP, which achieves through 80% accuracy. Based on result from the simulation used to display the train use test with cross validation is 10 fold in MLP is 29 for 63% in correctly classified instances (true positive rate for graduate) and 17 for 36% in incorrectly classified instances (false negative rate for quit).

Item Type: Undergraduates Project Papers
Additional Information: Project Paper (Bachelors of Computer Science (Software Engineering)) -- Universiti Malaysia Pahang – 2019, SV: DR. AHMAD FIRDAUS BIN ZAINAL ABIDIN, e-Thesis
Uncontrolled Keywords: Pharmaceutical student; machine learning
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Faculty/Division: Faculty of Computer System And Software Engineering
Depositing User: Mrs. Sufarini Mohd Sudin
Date Deposited: 11 Nov 2019 03:27
Last Modified: 20 Mar 2023 03:31
URI: http://umpir.ump.edu.my/id/eprint/26401
Download Statistic: View Download Statistics

Actions (login required)

View Item View Item