Machine learning classification model for identifying internet addiction among university students
In this era of globalization, Internet addiction is a concerning issue, especially among university students as they are required to use the internet for academic purposes. However, things might go wrong when they are addicted to the Internet as the Internet does not only provide knowledge but also...
Saved in:
Main Authors: | , , |
---|---|
Format: | Proceeding Paper |
Language: | English English |
Published: |
IEEE
2023
|
Subjects: | |
Online Access: | http://irep.iium.edu.my/109354/1/109354_Machine%20learning%20classification%20model.pdf http://irep.iium.edu.my/109354/7/109354_Machine%20learning%20classification%20model_SCOPUS.pdf http://irep.iium.edu.my/109354/ https://ieeexplore.ieee.org/abstract/document/10361435 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | In this era of globalization, Internet addiction is a concerning issue, especially among university students as they are required to use the internet for academic purposes. However, things might go wrong when they are addicted to the Internet as the Internet does not only provide knowledge but also entertainment such as music, videos, games, social media, etc. Internet addiction was exposed to the public when Young introduced Internet addiction in her study as well as an assessment for Internet addiction known as Young's Internet addiction test (IAT) which is a questionnaire. Nonetheless, there are some issues associated with the questionnaire regarding the integrity and literacy of the participants as well as the experience of the specialist which might introduce inconsistencies in the assessment of one's Internet addiction level. Hence, the machine learning algorithm is introduced to replace the conventional assessment method for Internet addiction. In this study, three machine learning models are developed and compared. The three models include convolutional neural network (CNN), K-nearest neighbours (KNN), and logistic regression (LR). The low Alpha power band of the EEG data is transformed into spectrograms and utilized as the input for the machine learning models. The spectrograms are presented as images and fed into the CNN model. On the other hand, as KNN and LR could not take in images as the input data, the magnitude of each frequency in every time segment of each spectrogram is computed and fed into the KNN and LR. The results show that CNN gives the best performance in terms of overall accuracy, precision, recall, and F1-score, while KNN gives the most consistent performance. |
---|