Article Per Year (5 Year)
p-Index From 2020 - 2025
1.931
P-Index
This Author published in this journals
All Journal JOIV : International Journal on Informatics Visualization Journal of Applied Data Sciences International Journal of Advances in Artificial Intelligence and Machine Learning
Zakaria, Mohd Zaki
Unknown Affiliation
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management

Published : 13 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 13 Documents
Search

 1   2 
Clustering the Unlabeled Data Using a Modified Cat Swarm Optimization Dewi, Deshinta Arrova; Kurniawan, Tri Basuki; Zakaria, Mohd Zaki; Armoogum, Sheeba
Journal of Applied Data Sciences Vol 5, No 3: SEPTEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i3.349

Abstract

This paper presents a modified version of the Cat Swarm Optimization (CSO) algorithm aimed at addressing the limitations of traditional clustering methods in handling complex, high-dimensional datasets. The primary objective of this research is to improve clustering accuracy and stability by eliminating the mixture ratio (MR), setting the counts of dimensions to change (CDC) to 100%, and incorporating a new search equation in the tracing mode of the CSO algorithm. To evaluate the performance of the modified algorithm, five classic datasets from the UCI Machine Learning Repository—namely Iris, Cancer, Glass, Wine, and Contraceptive Method Choice (CMC)—were used. The proposed algorithm was compared against K-Means and the original CSO. Performance metrics such as intra-cluster distance, standard deviation, and F- measure were used to assess the quality of clustering. The results demonstrated that the modified CSO consistently outperformed the competing algorithms. For example, on the Iris dataset, the modified CSO achieved a best intra-cluster distance of 96.78 and an F-measure of 0.786, compared to 97.12 and 0.781 for K-Means. Similarly, for the Wine dataset, the modified CSO reached a best intra-cluster distance of 16399, surpassing K-Means which recorded 16768. In conclusion, the modifications introduced to the CSO algorithm significantly enhance its clustering performance across diverse datasets, producing tighter and more accurate clusters with improved stability. These findings suggest that the modified CSO is a robust and effective tool for data clustering tasks, particularly in high-dimensional spaces. Future work will focus on dynamic parameter tuning and testing the scalability of the algorithm on larger and more complex datasets.
Gum Disease Identification Using Fuzzy Expert System Nasir, Muhammad; Kurniawan, Tri Basuki; Dewi, Deshinta Arrova; Zakaria, Mohd Zaki; Bujang, Nurul Shaira Binti
Journal of Applied Data Sciences Vol 5, No 3: SEPTEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i3.346

Abstract

Gum disease, including Gingivitis and Periodontitis, is among the most common dental conditions, primarily caused by dental plaque, a bacterial biofilm. These conditions are strongly linked to various systemic illnesses, including cancer, atherosclerosis, hypertension, stroke, and respiratory and cardiovascular conditions like aspiration pneumonia, as well as adverse pregnancy outcomes. Gum inflammation is typically characterized by symptoms such as increased redness, swelling (edema), and a loss of surface texture (stippling; gum fiber attachment). These symptoms are site-specific, meaning that an individual can have both healthy and diseased areas within their mouth. In this research, we developed a fuzzy expert system using MATLAB to identify gum diseases. The system was tested on various cases and produced an output value of 0.133, which successfully identified Gingivitis. This value was derived using a fuzzy logic system that processes input data through predefined rules within the Fuzzy Expert System (FES). The system utilizes several input variables such as the frequency of gum bleeding, the extent of plaque accumulation, the depth of gum recession, and the degree of tooth mobility. The key contribution of this study lies in the integration of fuzzy logic to handle the inherent uncertainties in clinical diagnosis, providing a more nuanced assessment compared to traditional methods. The novelty of this research is the application of a fuzzy expert system in dental diagnostics, offering a promising tool for improving the accuracy and efficiency of gum disease identification in clinical settings. This system has the potential to assist dentists in making more informed decisions, ultimately leading to better patient outcomes.
Fake vs Real Image Detection Using Deep Learning Algorithm Fatoni, Fatoni; Kurniawan, Tri Basuki; Dewi, Deshinta Arrova; Zakaria, Mohd Zaki; Muhayeddin, Abdul Muniif Mohd
Journal of Applied Data Sciences Vol 6, No 1: JANUARY 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i1.490

Abstract

The purpose of this research project is to address the growing issues presented by modified visual information by developing a deep learning model for identifying between real and fake images. To enhance accuracy, this project evaluates the effectiveness of deep learning algorithms such as Residual Neural Network (ResNet), Visual Geometry Group 16 (VGG16), and Convolutional Neural Network (CNN) together with Error Level Analysis (ELA) as preprocessing the dataset. The CASIA dataset contains 7,492 real images and 5,124 fake images. The images included are from a wide range of random subjects, including buildings, fruits, animals, and more, providing a comprehensive dataset for model training and validation. This research examined models' effectiveness through experiments, measuring their training and validation accuracies. It comes out with the best accuracy of each model, which is for Convolutional Neural Network (CNN), 94% for training accuracy, and validation accuracy of 92%. For VGG16, with both training and validation accuracy reaching 94%. Lastly, Residual Neural Network (ResNet) demonstrated optimal performance with 95% training accuracy and 93% validation accuracy. This project also constructs a system prototype for practical applications, offering an interface for real-world testing. When integrating into the system prototype, only Residual Neural Network (ResNet) shows consistency and effectiveness when predicting both fake and real images, and this led to the decision to choose ResNet for integration into the system. Furthermore, the project identified several areas for improvement. Firstly, expanding the model comparison for discovering more successful algorithms. Next, improving the dataset preprocessing phase by incorporating filtering or denoising techniques. Lastly, refining the system prototype for greater appeal and user-friendliness has the potential to attract a larger audience.
Deep Learning Based Face Mask Detection System Using MobileNetV2 for Enhanced Health Protocol Compliance Fadly, Fadly; Kurniawan, Tri Basuki; Dewi, Deshinta Arrova; Zakaria, Mohd Zaki; Hisham, Putri Aisha Athira binti
Journal of Applied Data Sciences Vol 5, No 4: DECEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i4.476

Abstract

Personal protective equipment (PPE) is crucial in mitigating the spread of infections within the pharmacy industry, manufacturing sectors, and healthcare facilities. Airborne particles and contaminants can be released during the handling of pharmaceuticals, the operation of machinery, or patient care activities. These particles can be transmitted through close contact with an infected individual or by touching contaminated surfaces and then touching one's face (mouth, nose, or eyes). PPE, including face masks, plays a vital role in minimizing the risk of transmission of infectious diseases. Although mandates for wearing face masks might relax as situations improve and vaccination rates increase, staying prepared for potential future outbreaks and the resurgence of infectious diseases remains important. Therefore, an automated system for face mask detection is important for future use. This research proposes real-time face mask detection by identifying who is (i) not wearing a mask and (ii) wearing a mask. This research presents a deep-learning approach using a pre-trained model, MobileNet-V2. The model is trained on a 10,000 dataset of images of individuals with and without masks. The result shows that the pre-trained MobileNet-V2 model obtained a high accuracy of 98.69% on the testing dataset.
Deep Learning Incorporated with Augmented Reality Application for Watch Try-On Andri, Andri; Kurniawan, Tri Basuki; Dewi, Deshinta Arrova; Alqudah, Mashal Kasem; Alqudah, Musab Kasim; Zakaria, Mohd Zaki; Hisham, Putri Aisha Athira binti
Journal of Applied Data Sciences Vol 6, No 1: JANUARY 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i1.529

Abstract

In evaluating the dynamic landscape of online shopping, the integration of Augmented Reality (AR) technologies has emerged as a transformative force, redefining the way consumers engage with products in virtual environments. This research project investigates the intersection of deep learning and AR in the context of online shopping, with a particular focus on a Watch Try-On application. The experimentation involves the use of SSD MobileNet's models for real-time object detection aimed at enhancing the user experience during online watch shopping. Training both SSD MobileNet's V1 and V2 models through 50,000 iterations, the results reveal intriguing insights into their performance. SSD MobileNet's V1 demonstrated superior results, boasting a mean average precision (mAP) of 0.9725 and a significant reduction in total loss from 0.774 to 0.5405. However, the longer training time of 7 hours and 42 minutes prompted the selection of SSD MobileNet's V2 for real-time applications due to its faster inference capabilities. Extending beyond traditional online shopping experiences, the research explores the potential of AR technologies to revolutionize product visualization and interaction. The choice of the Vuforia model target for the Watch Try-On application showcases the synergy between deep learning and AR, allowing users to virtually try on watches and visualize them in their real-world environment. The application successfully detects users' hands with high accuracy, creating an immersive and visually enriching experience. In conclusion, this project contributes to the ongoing discourse on the fusion of deep learning and AR for online shopping. The exploration of SSD MobileNet's models, coupled with the integration of AR technologies, underscores the potential to elevate the online shopping experience by providing users with dynamic, interactive, and personalized ways to engage with products.
Leveraging Data Analytics for Student Grade Prediction: A Comparative Study of Data Features Misinem, Misinem; Kurniawan, Tri Basuki; Dewi, Deshinta Arrova; Zakaria, Mohd Zaki; Nazmi, Che Mohd Alif
Journal of Applied Data Sciences Vol 5, No 4: DECEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i4.442

Abstract

In educational settings, a persistent challenge lies in accurately identifying and supporting students at risk of underperformance or grade retention. Traditional approaches often fall short by applying generalized interventions that fail to address specific academic needs, leading to ineffective outcomes and increased grade repetition. This study advocates for integrating machine learning algorithms into educational assessment practices to address these limitations. By leveraging historical and current performance data, machine learning models can help identify students needing additional support early in their academic journey, allowing for precise and timely interventions. This research examines the effectiveness of three machine learning algorithms: Naive Bayes, Deep Learning, and Decision Trees. Naive Bayes, known for its simplicity and efficiency, is well-suited for initial data screening. Deep Learning excels at uncovering complex patterns in large datasets, making it ideal for nuanced predictions. Decision Trees, with their interpretable and actionable outputs, provide clear decision paths, making them particularly advantageous for educational applications. Among the models tested, the Decision Tree algorithm demonstrated the highest performance, achieving an accuracy rate of 86.68%. This high precision underscores its suitability for educational contexts where decisions need to be based on reliable, interpretable data. The results strongly support the broader application of Decision Tree analysis in educational practices. By implementing this model, educational administrators can better identify at-risk students, tailor interventions to meet individual needs, and ultimately improve student success rates. This study suggests that Decision Trees could become a vital tool in data-driven strategies to enhance student retention and optimize academic outcomes.
Recommender System for Book Review based on Clustering Algorithms Udariansyah, Devi; Kurniawan, Tri Basuki; Dewi, Deshinta Arrova; Zakaria, Mohd Zaki; Hanan, Nur Syuhana binti Abd
Journal of Applied Data Sciences Vol 6, No 1: JANUARY 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i1.492

Abstract

Book reviews show the expression of the reviewers that are to be evaluated and describe the book. Today, the amount of the book is growing rapidly, and it offers people a lot of choices. The recommender system on book reviews is mostly mentioned, and we will recommend a book based on the keyword selected. This study highlights two primary objectives. The first objective is to identify the keywords of the book review, and the last objective is to design and develop a book review analysis visualization using the result of the k-means clustering algorithm. The methodology of this research consists of ten phases, which start with the preliminary study, knowledge acquisition and analysis phase, data collection phase, data pre-processing phase, and modeling phase. The research then continues with the design and implementation, dashboard development, testing and evaluation, and finally, the documentation phase. The data from this study is scraped from Amazon.com and focuses on three genres: Fiction and Fantasy, Mystery and Thriller, and Romance. All the data will be clean before it can be applied to k-means clustering. The result of clustering will define the keywords for every genre and will compare with the keywords for each book that was collected from Amazon.com.
Machine Learning Techniques for Distinguishing Android Malware Variants Irwansyah, Irwansyah; Kurniawan, Tri Basuki; Dewi, Deshinta Arrova; Zakaria, Mohd Zaki; Azmi, Nurhafifi Binti
Journal of Applied Data Sciences Vol 6, No 1: JANUARY 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i1.493

Abstract

The advancement of portable devices has been quickly and dramatically reshaping the usage trend and consumer preferences of electronic devices. Android, the most common mobile operating system, has a privilege-separated protection system with a complex access control mechanism. Android apps require permission to get access to confidential personal data and device resources. However, studies have shown that various malicious applications can acquire permission and target systems and applications by misleading users. In this study, we suggest a machine-learning approach to classifying Android malware variants by mining requested permissions, real permissions, suspicious calls, and API calls that were obtained and used in Android malware applications. Selected features were selected using a feature selection called KBest. Feature selection techniques are used to minimize the scale of the features and increase the performance. Two types of Naïve Bayes classifiers, called Multinomial distribution and multivariate Bernoulli distribution, are used and compared in malware family classification for text classification. Both naïve Bayes types are evaluated using a confusion matrix based on 4022 Android malware applications belonging to 10 families. Experimental findings show that the Multinomial distribution offers a reliable performance from three tests experiment with an average accuracy of 95%.
Music Recommendation Based on Facial Expression Using Deep Learning Kurniawan, -; Kurniawan, Tri Basuki; Dewi, Deshinta Arrova; Zakaria, Mohd Zaki; Saringat, Zainuri; Firosha, Ardian
JOIV : International Journal on Informatics Visualization Vol 9, No 1 (2025)
Publisher : Society of Visual Informatics

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62527/joiv.9.1.3794

Abstract

Music's profound impact on human emotions is essential for creating personalized experiences in entertainment and therapeutic settings. This study introduces a cutting-edge music recommendation system that utilizes facial expression analysis to tailor music suggestions according to the user's emotional state. Our approach integrates a haar-cascade classifier for real-time face detection with a Convolutional Neural Network (CNN) that classifies emotions into seven distinct categories: happiness, sadness, anger, fear, disgust, surprise, and neutrality. This emotionally aware system recommends music tracks corresponding to the user's current emotional condition to enhance mood regulation and overall listener satisfaction. The effectiveness of our system was evaluated through rigorous testing, where the CNN model demonstrated a high degree of accuracy. Notably, the model achieved an overall accuracy of 84.44% in recognizing facial expressions. Precision, recall, and F1 scores consistently exceeded 84%, indicating robust performance across diverse emotional states. These results underscore the system's capability to accurately interpret and respond to complex emotional cues through tailored music suggestions. Integrating advanced deep learning techniques for face and emotion recognition enables our recommendation system to adapt dynamically to the user's emotional fluctuations. This responsiveness ensures a highly personalized music listening experience that reflects the user's feelings and potentially enhances their emotional well-being. By bridging the gap between static user profiles and the dynamic nature of human emotions, our system sets a new standard for personalized technology in music recommendation, promising significant improvements in user engagement and satisfaction.
Detecting Gender-Based Violence Discourse Using Deep Learning: A CNN-LSTM Hybrid Model Approach Kurniawan, Tri Basuki; Dewi, Deshinta Arrova; Henderi, Henderi; Hasibuan, M. Said; Zakaria, Mohd Zaki; Ismail, Abdul Azim Bin
Journal of Applied Data Sciences Vol 6, No 3: September 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i3.761

Abstract

Gender-Based Violence (GBV) is a critical social issue impacting millions worldwide. Social media discussions offer valuable insights into public awareness, sentiment, and advocacy, yet manually analyzing such vast textual data is highly challenging. Traditional text classification methods often struggle with contextual understanding and multi-class categorization, making it difficult to accurately identify discussions on Sexual Violence, Physical Violence, and other topics. To address this, the present study proposes a hybrid deep learning approach combining Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks. CNN is utilized for extracting key linguistic features, while LSTM enhances the classification process by maintaining sequential dependencies. This hybrid CNN+LSTM model is evaluated against standalone CNN and LSTM models to assess its performance in classifying GBV-related tweets. The dataset was sourced from Kaggle, containing real-world Twitter discussions on GBV. Experimental results demonstrate that the hybrid model surpasses both CNN and LSTM models, achieving an accuracy of 89.6%, precision of 88.4%, recall of 89.1%, and F1-score of 88.7%. Confusion matrix and ROC curve analyses further confirm the hybrid model’s superior performance, correctly identifying Sexual Violence (82%), Physical Violence (15%), and Other (3%) cases with reduced misclassification rates. These results suggest that combining CNN’s feature extraction with LSTM’s contextual learning provides a more balanced and effective classification model for GBV-related text. This work supports the development of AI-based tools for social media monitoring, policy-making, and advocacy, helping stakeholders better understand and respond to GBV discussions. Future research could explore transformer-based models like BERT and real-time classification applications to further improve performance.
Co-Authors - Kurniawan, - Alqudah, Mashal Kasem Alqudah, Musab Kasim Andri Andri Armoogum, Sheeba Azmi, Nurhafifi Binti Bujang, Nurul Shaira Binti Dewi, Deshinta Arrova Dhidhi Pambudi Efrizoni, Lusiana Eka, Eka Puji Agustini Fadly Fadly Fatoni, Fatoni Firosha, Ardian Hanan, Nur Syuhana binti Abd Henderi . Hisham, Putri Aisha Athira binti Irwansyah Irwansyah Ismail, Abdul Azim Bin Kasem Alqudah, Mashal Kurniawan, Tri Basuki M Said Hasibuan Misinem, Misinem Muhammad Nasir Muhayeddin, Abdul Muniif Mohd Nazmi, Che Mohd Alif Saringat, Zainuri Shahreen Kasim, Shahreen Udariansyah, Devi