Journal of Telecommunication, Electronic and Computer Engineering (JTEC)

Malaysian Community College Graduates Employability Prediction Model Using Machine Learning Approach

Azida Mansor — 2024-09-30

Community College, as TVET institution under the Ministry of Higher Education, offers industry-relevant skills training to ensure graduates' employability in the global labor market. However, producing graduates who meet industry demands remains a challenge, and industries continue to face difficulties in obtaining skilled graduates. In Malaysia, there is limited research on predictive models for employability rates among graduates of TVET Malaysian institutions. This research utilizes Python to investigate the significant factors influencing employability among Malaysian Community College graduates, determined by a specific indicator of whether they will be employed. Our contribution lies in developing an accurate employability prediction model using machine learning algorithms such as Logistic Regression, Neural Networks, and Random Forest. The dataset used consisted of 10,427 instances and 14 attributes, from which six significant factors were identified. Among the models, Random Forest outperformed the other machine learning models, and hyperparameter tuning using RandomizedSearch further improved the accuracy of the model to 84.8%. This study aims to identify the most accurate and interpretable model, providing valuable insights for educational institutions to enhance their employability strategies.

Impact Analysis of Filter and Wrapper-Based Feature Selection Techniques for Webpages Phishing Attacks Identification

Patrick Olabisi — 2024-09-30

Phishing, which involves fraudulently gaining access to sensitive assets of unsuspecting individuals through deceptive and malicious emails, is a major global threat to internet users. The proliferation of phishing sites and their operations is occurring at an alarming rate, raising significant concerns about how to forestall them. Numerous research efforts are underway to detect phishing attempts before they can compromise important information and cause damage. Compared to conventional methods, machine learning has proven highly effective at detecting phishing attacks by analyzing different features. This study analyzed the behaviors of seven classification data mining algorithms on optimal subset features selection using Wrapper (Boruta) and Filter-based (Mutual-Information). Real-life phishing webpage datasets were used for the analysis. Ensemble classifiers such as Voting, Gradient Boosting, and Random Forest were used in the experiments. Two experiments were conducted. In the first experiment, K-Nearest Neighbor (K-NN) yielded the highest accuracy among single classifiers, with a score of 94.1%, while Random Forest (RF) ensemble achieved 96.7%. In the second experiment, using another baseline feature set, RF performed excellently under the Boruta method with an accuracy of 97.25%, while K-NN retained the highest accuracy of 95.20% among single classifiers. This study provides empirical evidence that feature selection techniques have a great impact on the performance of ML models, for both single and ensemble classifiers, in the detection of phishing attacks.

Kurdish Sign Language Recognition Using Convolutional Neural Network (CNN)

Sarkhel H. Taher Karim — 2024-09-30

The present study examines the obstacles encountered by the deaf population, with a specific emphasis on the growing importance of sign language in facilitating effective communication. The main mode of communication for deaf individuals is Sign Language (SL), which conveys meaning visually and expressively through facial expressions, hand movements, and body gestures. The objective of this project is to automate the recognition of sign language to improve accessibility and reduce reliance on interpreters. Specifically, this work focused on developing an alphabet recognition system for Kurdish Sign Language (KSL). Due to its many intricacies and resemblances to the Arabic script, KSL requires a robust recognition model. The proposed method utilizes Convolutional Neural Networks (CNN) trained on a real-world dataset to accurately recognize both numerical values and alphabetic characters in the Kurdish Sign Language (KSL). The real-time operation of the system enables rapid recognition of hand gestures, providing immediate textual output. The dataset used for training comprises 132,000 hand images, including 33 alphabetic signs and numeral signs from 0 to 9. The use of MediaPipe, a method for processing 3D images, significantly improves the efficiency of gesture detection. Multiple methodologies were investigated, and the integration of Convolutional Neural Networks (CNN), TensorFlow, and MediaPipe resulted in a remarkable accuracy of 99.87% with negligible dropout rates. This study establishes a foundation for enhanced communication and independence for the deaf community, representing a significance advancement in the automation of sign language recognition.

Design and Development of Lightning Detection System Utilizing Slow Atmospheric Electric Field Waveform at Legoland Malaysia

Erman Ramli — 2024-09-30

Conventional lightning localization and detection techniques, including Magnetic Direction Finder (MDF), Time of Arrival (TOA), Interferometer (ITF), and Distance of Arrival (DOA), predominantly rely on fast atmospheric electric fields, magnetic fields, and very high frequency (VHF) signals. This paper pioneers a novel approach by delving into the analysis of the slow atmospheric electric field, aiming to develop a waveform analysis utilizing this field to estimate the distance and radius of lightning occurrences at LEGOLAND Malaysia Resort. The newly implemented lightning detection system at LEGOLAND leverages a straightforward and cost-effective setup, incorporating a capacitive antenna, slow and fast atmospheric electric field sensors, and dedicated data analysis software. The system's efficacy and accuracy have undergone a rigorous comparison with LEGOLAND's existing online lightning detection service. Achieving accurate data necessitates proper grounding and isolation from electrical noise, as signal interference from power lines, towers, or machinery can potentially trigger false signals. This research has contributed detailed documentation on the analysis of slow atmospheric electric field data, encompassing waveform patterns and key characteristics. This documentation is expected to serve as a valuable resource for future research endeavors and the continuous refinement of lightning detection systems. The comparative evaluation between the novel system and the current online service at LEGOLAND Malaysia Resort has shed light on the efficiency and capabilities of the newly introduced methodology.

Enhanced Malaysian License Plate Recognition System Using an Improved YOLOv2 Model

Syafeeza A.R. — 2024-09-30

License Plate Recognition (LPR) has gained popularity among researchers due to its wide range of applications, including law enforcement, monitoring, and toll gate systems. However, existing LPR systems still require improvements to achieve optimum accuracy and speed. The advancements in Convolutional Neural Network (CNN) variants offer potential solutions for these challenges. This primary aim of this system is to ensure accurate and efficient recognition of the vehicle plate characters using CNN techniques. This research utilizes two CNN network architectures for deep object detection to address the Malaysian License Plate Recognition (MLPR) task. The first network is designed to detect the license plate, while the second is responsible for recognizing the characters on the plate. Both networks are cascaded from the architecture of two-stage YOLOv2, providing promising speed and accuracy. The MLPR system achieved an accuracy of 98.75% and a processing speed of 0.0104 seconds, using a total of 2,200 license plate images. In conclusion, the system adapted from deep object detection techniques presents a promising solution for the MLPR problem, based on the achieved accuracy and speed.