cover
Contact Name
Mustakim
Contact Email
Mustakim
Phone
-
Journal Mail Official
ijaidm@uin-suska.ac.id
Editorial Address
-
Location
Kab. kampar,
Riau
INDONESIA
Indonesian Journal of Artificial Intelligence and Data Mining
ISSN : 26143372     EISSN : 26146150     DOI : -
Core Subject : Science,
Indonesian Journal of Artificial Intelligence and Data Mining (IJAIDM) is an electronic periodical publication published by Puzzle Research Data Technology (Predatech) Faculty of Science and Technology UIN Sultan Syarif Kasim Riau, Indonesia. IJAIDM provides online media to publish scientific articles from research in the field of Artificial Intelligence and Data Mining. IJAIDM will be published 2 (two) times a year, in March and September, each edition contains 7 (seven) articles. Articles may be written in English or Indonesia.
Arjuna Subject : -
Articles 233 Documents
A Robot Model for Detecting Smoking Violations Using YOLOv5 and PID-Based Navigation Control Muslimin, Selamat; Megaarta, Muhammad Andaru; Triandika, Rayhan
Indonesian Journal of Artificial Intelligence and Data Mining Vol 8, No 2 (2025): July 2025
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24014/ijaidm.v8i2.37345

Abstract

Smoking violations in restricted areas, especially in public spaces exposed to secondhand smoke, remain a significant concern. This study develops an autonomous robot designed to detect smoking violations using YOLOv5 and Raspberry Pi. The robot's camera captures real-time images to identify smoking behavior, with YOLOv5 accurately detecting cigarette objects. For navigation, the robot employs a PID control system, complemented by an encoder and a compass sensor, ensuring precise movement. The results demonstrate that the robot achieves a confidence level of 87% in detecting smoking behavior at a distance of 250 cm, with a frame rate of 8 FPS. The PID-based navigation system ensures minimal error of ±5 cm over a 2-meter distance. These findings emphasize the robot's effectiveness in both detecting smoking violations and navigating accurately, making it an effective tool for the enforcement of smoke-free zone regulations.
Forex Price Predictions using Hybrid TCN-LSTM and LSTM-TCN Models Caroline, Caroline; Lestari, Wulan Sri; Ulina, Mustika
Indonesian Journal of Artificial Intelligence and Data Mining Vol 8, No 3 (2025): November 2025
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24014/ijaidm.v8i3.38354

Abstract

Forecasting financial market prices, particularly foreign exchange (forex) rates, remains a substantial difficulty due to the market's inherent unpredictability, intricacy, and turbulent characteristics. By combining the Temporal Convolutional Network (TCN) and Long Short-Term Memory (LSTM) models into a hybrid framework, this study overcomes this difficulty and improves prediction accuracy.  The MinMaxScaler function was used to standardize the input data prior to training, bringing all values into a range between 0 and 1.  An 80% training segment and a 20% testing segment were then separated from the prepared dataset.  We tested two different hybrid architectures, the LSTM-TCN and the TCN-LSTM, with the EUR/USD, AUD/USD, and GBP/USD value pairs.  With uniform parameters applied to both models during training, the Root Mean Squared Error (RMSE) measure was used for all performance evaluations in order to ensure a fair comparison and determine which model was better. The LSTM-TCN architecture proved to be the superior predictor on the testing set. It recorded a lower average RMSE of 0.003911. This result contrasts with the TCN-LSTM model's performance, which yielded a higher average RMSE of 0.004181.
Spice Image Classification Using ResNet50 and Augmentation Technique Bacun, Julio Francisco; Pratama, Irfan
Indonesian Journal of Artificial Intelligence and Data Mining Vol 8, No 3 (2025): November 2025
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24014/ijaidm.v8i3.37862

Abstract

This research aimed to develop an automatic classification system for Indonesian spices using a deep learning approach based on the ResNet50 architecture. The classification task involved 31 spice categories with 210 images per class. Two training strategies were implemented: training the model from scratch and using transfer learning with pre-trained weights from ImageNet. The model trained from scratch achieved a validation accuracy of 57%, while the transfer learning approach combined with fine-tuning of the last 33 layers resulted in a significantly higher validation accuracy of 96%. Image preprocessing, data augmentation, and class weighting were applied to improve the model’s generalization and handle data imbalance. The confusion matrix analysis showed that most predictions aligned with the true labels, especially in the transfer learning model. These findings demonstrate that transfer learning with ResNet50 can effectively classify spice images with high accuracy, even when visual similarity between certain classes exists. This research highlights the potential of deep convolutional neural networks to support automatic and reliable identification systems for biodiversity mapping and agricultural industries
Real-Time Detection of Autistic Children's Activities Using YOLOv8 on Social Monitoring Robots Prihatini, Ekawati; Muslimin, Selamat; Hadi, Kurnia
Indonesian Journal of Artificial Intelligence and Data Mining Vol 8, No 2 (2025): July 2025
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24014/ijaidm.v8i2.37380

Abstract

Children with autism spectrum disorder require special attention in both therapy and daily activity monitoring. One approach that can assist is the utilization of a Social Monitoring Robot (SMR) with the capability of automatic activity monitoring. This study aims to develop a real-time activity detection system for children with autism using the You Only Look Once version 8 (YOLOv8) algorithm on the SAR platform. The system is designed to recognize key activities such as eating, studying, and walking, through video input from a webcam processed by a Raspberry Pi. The recognition process is carried out by detecting bounding boxes and confidence scores for the child and their activities. The detection results are then visualized through a Human Machine Interface (HMI). Based on the testing, the system is capable of detecting and classifying children's activities with a fairly high level of reliability under real-world environmental conditions. These results indicate that the implementation of YOLOv8 in an SMR-based monitoring system has the potential to enhance supervision and intervention for children with autism in a more responsive and personalized manner.
Swin Transformer V2 for Invasive Ductal Carcinoma Classification in Histopathological Imaging Ariyanto, Puguh Aiman; Wisesty, Untari Novia
Indonesian Journal of Artificial Intelligence and Data Mining Vol 8, No 3 (2025): November 2025
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24014/ijaidm.v8i3.38606

Abstract

Breast cancer is the second leading cause of mortality in women globally, with Invasive Ductal Carcinoma being the most dominant subtype that requires accurate diagnosis to increase patient life expectancy. Conventional diagnosis based on manual histopathological examinations is time-consuming, prone to misinterpretation, and exhibits significant inter-observer variability. This study implemented the Swin Transformer V2 architecture for the automatic classification of Invasive Ductal Carcinoma on 277,524 histopathological images, each measuring 50×50 pixels, which were resized to 256×256 pixels with geometric augmentation. The model was trained using AdamW optimization with a learning rate of 1 × 10⁻⁴, weight decay of 1 × 10⁻⁴, a batch size of 16, and mixed precision (FP16) for five epochs at a 70:20:10 data sharing ratio. The data augmentation includes a 50% probability of a random horizontal flip and a maximum of 10 degrees of random rotation to improve the model's generalization capabilities. Evaluation of 27,754 independent test samples resulted in an accuracy of 92.82%, an accuracy of 88.48%, a recall of 86.05%, an F1-score of 87.25%, and an AUC of 0.91. A hierarchical window attention-shifted mechanism with residual post-normalization has been shown to be effective in extracting local and global features from complex microscopic images. The results show that Swin Transformer V2 has significant potential as a diagnostic aid system to enhance the efficiency and accuracy of early breast cancer detection in clinical pathology practice.
Implementation of an RFID RC522 and IoT-Based Automatic Door Security System in an Electrical Engineering Laboratory Wijanarko, Yudi; Alfarizal, Niksen; Pratama, Muhammad Regi
Indonesian Journal of Artificial Intelligence and Data Mining Vol 8, No 2 (2025): July 2025
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24014/ijaidm.v8i2.37007

Abstract

The development of smart home technology has encouraged the development of more efficient and integrated automatic security systems, Using RFID technology with the Internet of Things (IoT) is one method that can be applied. This research aims to implement an automatic door security system based on RFID RC522 connected to the IoT network in the Electrical Engineering Laboratory. The research method used is Research and Development (RnD) through the process of designing, making, and testing system prototypes. The system is controlled by an ESP32 microcontroller, uses an RC522 RFID module for authentication, and utilizes relays and solenoid door locks as locking mechanisms. Test results show that the maximum effective RFID reading distance is 5 cm, with a fast and accurate response at a distance of 1 to 4 cm. The system can log all access activities directly through the IoT platform and distinguish between valid and invalid cards. In terms of power consumption, the door lock solenoid has the largest power usage when active at 7.2 W, while other components remain efficient. The implementation of this system is proven to be able to improve the security aspects and ease of access monitoring in the laboratory room automatically and connected.
Field-Level AES-128 Encryption in Laravel-based E-Commerce for MSME Data Protection Afifah, Luthfia; Nurdin, Ali; Handayani, Ade Silvia
Indonesian Journal of Artificial Intelligence and Data Mining Vol 8, No 2 (2025): July 2025
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24014/ijaidm.v8i2.37675

Abstract

The increasing digitization of micro, small, and medium enterprises (MSMEs) in e-commerce brings critical challenges in protecting customer data. Despite the widespread use of encrypted communication protocols such as HTTPS and TLS for secure data transmission, many MSMEs still fail to implement encryption at the data storage level. This means that once the data reaches the server, it is often stored in unencrypted form within the database. This study implemented AES-128 encryption at the field-level in a Laravel-based e-commerce system to protect MSME customer data. The encryption was applied to sensitive data fields and tested through black-box testing and benchmark analysis. A dataset of 10,000 records was used to compare performance between plaintext and encrypted operations. Results showed an average encryption overhead of 0.0409 seconds, indicating minimal impact on performance. The encryption-decryption process consistently returned correct outputs across all trials. This solution offers an affordable and scalable encryption model for MSMEs, enhancing customer data security without relying on external tools or infrastructure.
Enhancing Student Performance Classification Through Dimensionality Reduction and Feature Selection in Machine Learning Mustakim, Mustakim; Sari, Windy Junita; Ulfa, Fara
Indonesian Journal of Artificial Intelligence and Data Mining Vol 8, No 3 (2025): November 2025
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24014/ijaidm.v8i3.37783

Abstract

Education plays an important role in shaping the intellectual and character of the nation's next generation. However, poor student academic performance is a major challenge, especially regarding student retention and dropout risk. This study aims to evaluate the performance of machine learning algorithms, namely K-Nearest Neighbor (K-NN), Light Gradient Boosting Machine (LightGBM), and Extreme Gradient Boosting (XGB), and analyze the effect of dimensionality reduction using Principal Component Analysis (PCA) and feature selection with Recursive Feature Elimination (RFE) on student performance prediction accuracy. The research dataset consists of 395 student samples with demographic, social, and academic attributes. The results show that XGB has the best performance with 98.32% accuracy and can predict all classes with perfect 100% accuracy. LightGBM and K-NN achieved 94.87% and 93.88% accuracy, respectively. The best attributes affecting student performance were found in the “Highly Prioritized” category, including study time, family support, family, and health. Although PCA slightly degraded the model performance, feature selection with RFE significantly improved accuracy. This study concludes that proper algorithm selection and focus on relevant attributes can improve prediction accuracy and efficiency, making an important contribution to the development of more effective education prediction systems.
Development of a Raspberry Pi 4-Powered Internet of Things System for Acne-Prone Skin Health Monitoring Kurniawan, Aprila; Sari, Dewi Permata; Wijanarko, Yudi; Sabara, Gally
Indonesian Journal of Artificial Intelligence and Data Mining Vol 8, No 2 (2025): July 2025
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24014/ijaidm.v8i2.36997

Abstract

This research developed an Internet of Things (IoT)-based facial skin health monitoring system, with a focus on acne-prone skin. Facial skin is categorized into three main types: normal, oily, and dry, as well as four types of acne: blackheads, papules, pustules, and nodules. The system is designed to enhance the accuracy of skin condition monitoring through facial image analysis, utilizing a dataset of 4,092 images. The high number of acne cases, especially in 12-24 year olds with 40-50 million cases in the United States, is the background of this research. Conventional skin analyzers are considered less capable of providing accurate quantitative data. Therefore, a Smart Skin Analyzer Detector was developed that uses a Raspberry Pi as a data processor. Images are taken through a webcam, analyzed, and then the results are sent to the cloud. The system is also integrated with Telegram to provide users with real-time notifications regarding their skin type and acne condition. This approach enables more effective, faster, and more affordable skin monitoring. The results demonstrate that IoT technology has significant potential in enhancing personalized and sustainable skin care.
Comparative Evaluation of Optuna-Optimized Radial Basis Function and Sigmoid Kernels in Support Vector Machine for Smart Air Quality Classification Galea, Nanda; Rahman, A; Maulidda, Renny; Husni, Nyayu Latifah
Indonesian Journal of Artificial Intelligence and Data Mining Vol 8, No 3 (2025): November 2025
Publisher : Universitas Islam Negeri Sultan Syarif Kasim Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24014/ijaidm.v8i3.37563

Abstract

Poor air quality can have a serious impact on human health, so a classification system capable of accurately identifying air conditions is needed. This research proposes an air quality classification method using the Support Vector Machine (SVM) algorithm with two types of non-linear kernels, namely Radial Basis Function (RBF) and Sigmoid. The data used is obtained from various environmental sensors that record parameters such as CO, smoke, HC, TVOC, eCO₂, temperature, and humidity, and then collected in the form of historical datasets. To enhance the accuracy and efficiency of the model, hyperparameter optimization was performed automatically using Optuna. The evaluation results showed that SVM with RBF kernel performed better than Sigmoid kernel, achieving an accuracy value of 96.67% and F1-score of 96.80%. In addition, RBF also showed higher stability in 5-fold cross validation. This research shows that the combination of SVM and Optuna is effective in building an accurate air quality classification system, and has the potential to be further developed as a sensor based in air monitoring system and IoT.