Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi)
Not a member yet
1071 research outputs found
Sort by
Face Dermatological Disorder Identification with YoloV5 Algorithm
Full text linkDermatological disorders are common in humans. The accurate identification of skin diseases is paramount for determining the most efficacious treatment. This system can screen images of skin diseases on the face and provide analysis results in the form of object detection. Dermatological disorders of the face are classified into six categories: acne nodules, melasma, filiform warts, milia, papules, and pustules. The YoloV5 algorithm was selected because of its effectiveness in live-detection tasks. The image-enhancement process involves the implementation of two methodologies: sharpening and histogram equalization. The former adjusts the brightness values whereas the latter adjusts the contrast values. The dataset comprised 1,223 images of skin diseases, with 947 images allocated for training and 276 for validation. The optimal mAP of the filiform wart class was determined to be 87.6%, with values of 76.7% for pustules, 72% for papules, 71% for milia, 68% for nodules, and 38.2% for melasma, representing the lowest value. The low mAP of melasma was attributed to the abstract image data type and complexity of localization. The congruence of object features and disparity in data variance has the potential to influence outcomes
A New Triple-Weighted K-Nearest Neighbor Algorithm for Tomato Maturity Classification
No full textAs climatic products, tomatoes are highly sensitive to harvesting and processing. The sorting of tomatoes can be significantly improved by utilizing Hue Saturation Value (HSV) color features that are classified using neighboring algorithms, such as K-Nearest Neighbor (KNN), Weighted K-Nearest Neighbor (W-KNN), and DW-KNN. However, the DW-KNN algorithm does not consider the relative relationship between the farthest, nearest, and surrounding neighbors, which may impact the classification accuracy, particularly in datasets with uneven distributions. This study proposes a Triple Weighted K-Nearest Neighbor (TW-KNN) algorithm for tomato image classification. This algorithm effectively handles the problem of sensitivity and outliers in the data distribution and considers the relationship between neighboring distances. The classification data consisted of 400 tomato images with five maturity levels divided into training and testing sets using k-fold cross-validation. Tests were conducted using several variations of parameter k, namely 4, 6, 9, and 15, to evaluate the classification performance. The results show that the proposed TW-KNN algorithm consistently outperforms other methods by producing better classification results. This is demonstrated by an accuracy rate of 95.52% across different values of k. The superior performance of the TW-KNN highlights its ability to provide robust and stable classification results compared to conventional KNN variants. This finding indicates that the TW-KNN is more effective in consistently classifying tomato fruits, making it a promising approach for automated fruit sorting applications
Hyperparameter Tuning with Optuna to optimize the YOLOv11n Model for Weed Detection
No full textAccurate weed detection is essential for maintaining the cleanliness and aesthetic appeal of residential yards. This study aimed to optimize YOLOv11n, a lightweight object detection model, to achieve high precision in weed identification under real-world conditions. The novelty of this study lies in the application of Optuna, an automatic hyperparameter optimization framework, to enhance model performance while maintaining computational efficiency for resource-limited devices such as drones and IoT systems. The research involved data augmentation techniques including crop (0–20% zoom), hue (±20°), saturation (±30%), brightness (±20%), exposure (±15%), and mosaic augmentation. These augmented images were used to train four YOLO nano variants (v5n, v8n, v11n, v12n), which were evaluated using standard metrics: Precision, Recall, F1-Score, and mean Average Precision (mAP). Among the models tested, YOLOv11n with Custom Optuna configuration delivered the highest performance, achieving a 94.6% F1-score and 97.8% [email protected]. These results demonstrate that the optimized YOLOv11n model can support accurate and efficient real-time weed detection in household environments, particularly on edge devices with limited hardware capabilities. This makes it a viable solution for practical implementation in precision agriculture and smart gardening
Performance Comparison of YOLOv8 and DETR in White Blood Cell Detection
No full textAutomated detection and classification of white blood cells (WBCs) from microscopic images play a vital role in supporting the diagnosis of hematological diseases. Accurate and robust object detection algorithms are essential for handling interclass similarities and imbalanced datasets. This study aims to evaluate and compare the performance of two modern object detection algorithms—Detection Transformer (DeTR) and YOLOv8—in performing multiclass WBC classification using public datasets from various sources with diverse visual characteristics. Five experimental scenarios were designed based on varying class distributions and data augmentation techniques, including horizontal/vertical flipping and random rotation. Both methods were trained and evaluated on the same dataset partitions, and their performances were assessed using the following standard metrics: precision, recall, and F1-score for each WBC class. The results show that YOLOv8 consistently achieved superior and more stable performance across all scenarios, with average F1-scores close to 1.00 even in augmented and imbalanced conditions. In contrast, DeTR performed competitively in balanced scenarios but showed lower consistency, particularly in classes such as Neutrophil and Monocyte. Data augmentation positively affected both models, although the gains were more prominent in YOLOv8. This study highlights the strong potential of YOLOv8 in real-time WBC classification tasks and presents DeTR as a viable yet less-optimized approach for this application. These findings contribute to the advancement of medical image-based object detection and offer valuable insights into the selection of appropriate algorithms for hematological image analysi
Explainable Ensemble Learning Framework with SMOTE, SHAP and LIME for Predicting 30-Day Readmission in Diabetic Patients
No full textHospital readmission among diabetic patients poses a significant burden on healthcare systems due to its frequency and associated costs. This study presents a machine learning framework for predicting 30-day readmission in diabetic patients using the Diabetes 130-US Hospitals dataset. The framework integrates data preprocessing, SMOTE for class balancing, ensemble learning, and explainable AI (SHAP and LIME) to enhance both accuracy and interpretability. Multiple models were evaluated, and the best performance was achieved by a weighted ensemble with a recall of 89.43% and an F1-score of 0.6612, indicating strong sensitivity. Explainability analysis using SHAP and LIME highlighted key predictors, notably Medication Change Status and Inpatient Admissions, which are clinically relevant. By combining predictive performance with transparent explanations, the proposed framework offers a practical and trustworthy tool for clinical decision support in managing diabetic readmissions
ResNet50-Driven Quality Inspection for Recorder Musical Instrument
No full textThe manufacturer of a recorder musical instrument requires high-quality product. The aim is to produce precise tones and an aesthetic look at customer satisfaction. A major challenge encountered by manufacturers is traditional visual inspection. Human error is a major factor, notably over extended work periods and the subjective judgment of quality control personnel. This paper reports on the development of a machine vision system for detecting abnormal patterns on the inner surface of a recorder musical instrument. An industrial-grade camera with a resolution of 1280 × 1024, paired with industrial lighting, was utilized. Due to its tube-shaped construction of the object, the bright-field imaging technique is applied to illuminate the interior. ResNet50 was selected as a feature extractor due to its balance between accuracy and efficiency. In addition, a Neural Network served as the classifier. A total of 1,118 images were collected as training data and 304 images as testing data. The training and testing data were separate sets that were taken independently, preventing any risk of data leakage. The test results indicated that the model performed exceptionally well in classification, achieving an accuracy of 95.7%, precision of 95.45%, sensitivity of 96.07%, and specificity of 95.36%. Moreover, the area under the curve of the Receiver Operating Characteristic (ROC AUC) score in test data reached 0.9906, reflecting the model's ability to separate features from the two classes. These findings suggest that the proposed method offers an alternative to subjective visual inspection. Future research may examine diverse deep learning architectures to further enhance performance while achieving faster classification
The Effect of Hyperparameters on Faster R-CNN in Face Recognition Systems
Full text linkFace recognition is one of the main challenges in the development of computer vision technology. This study aims to develop a face recognition system using a Faster R-CNN architecture, optimized through hyperparameter tuning. This research utilizes the "Face Recognition Dataset" from Kaggle, which comprises 2,564 face images across 31 classes. The development process involves creating bounding boxes using the LabelImg application and implementing the Grid Search method. The Grid Search is applied with predefined hyperparameter combinations (3 epochs [10, 25, and 50] × 3 learning rates [0.001, 0.0001, and 0.00001] × 3 optimizers [SGD, Adam, and RMS], resulting in 27 models). The evaluation metrics used were accuracy, precision, recall, and F1-score. The experimental results show that the selection of hyperparameters significantly affects the model performance. Based on the experimental results, the combination of the learning rate 0.00001, 50 epochs, and Adam optimizer yielded the highest accuracy and improvement of 8.33% compared to the baseline model. The results indicate that hyperparameter optimization enhances the ability of the model to recognize faces. Compared to conventional models, a Faster R-CNN performs better in detecting faces more accurately. Future research could further enhance the face recognition efficiency and accuracy by exploring other deep learning architectures and more advanced hyperparameter optimization techniques
Obesity Status Prediction Through Artificial Intelligence and Balanced Label Distribution Using SMOTE
Full text linkObesity, a global health challenge influenced by genetic and environmental factors, is characterized by excessive body fat that increases the risk of various diseases. With over two billion individuals affected worldwide, addressing this issue is crucial. This study investigated the application of Artificial Intelligence (AI) to predict obesity status using a dataset of 1,610 individuals, including demographic and anthropometric data. Four AI algorithms were analyzed: Artificial Neural Network (ANN), K-Nearest Neighbors (KNN), Random Forest, and Support Vector Machine (SVM). The Synthetic Minority Over-Sampling Technique (SMOTE) was applied to address dataset imbalance. The results demonstrate that SMOTE significantly enhanced the models' performance, especially in recall and F1-score for minority classes, such as obesity. Random Forest achieved the highest accuracy (92%) and recall (92%) post-SMOTE. The ANN showed substantial improvement in recall, increasing from 77% to 89%, whereas the SVM achieved the highest precision (89%), minimizing false positives. Despite these improvements, KNN remained the least effective. The findings underscore the critical role of SMOTE in improving AI model accuracy for obesity prediction and highlight Random Forest as the most reliable algorithm for clinical decision-making. Limitations, such as dataset representativeness, suggest future research directions, including expanding data diversity and advanced feature selection techniques. This study provides valuable insights into leveraging AI and preprocessing methods for obesity management.
 
Evaluating Steganography Detection in JPEG Images Using Gaussian Mixture Model and Cryptographic Keys
No full textThis study introduces a novel approach that integrates Gaussian Mixture Models (GMM) with MD5 hash-based verification to detect hidden messages embedded via Least Significant Bit (LSB) steganography in JPEG images. Unlike previous methods, the proposed dual-layer technique combines probabilistic modeling with data integrity verification. The model was trained and evaluated using a dataset comprising both original and stego-JPEG images. The experimental results achieved an accuracy of 78.67% and a precision of 89.15%, indicating good class separation between stego and non-stego images (AUC-ROC = 0.8659). However, the recall rate of 69.70% suggests that there is room for improvement in detecting all stego instances. Although MD5 is a hash function rather than an encryption algorithm, it effectively aids in identifying data anomalies resulting from message embedding. Overall, this lightweight approach offers a practical solution for steganalysis and can be further enhanced through the integration of hybrid deep learning techniques in future research
Evaluating the Accuracy of a Hybrid Neural Model with RBF-Polynomial Kernel for Rainfall Prediction: A Comparative Analysis of Trainlm and Trainrp Functions
No full textAccurate rainfall prediction is crucial for effective water management and disaster mitigation. This study introduces a novel hybrid neural model that employs a fourth-degree polynomial kernel and provides the first empirical comparison of the trainlm and trainrp functions to enhance forecasting accuracy. This study explored the application of a neural network algorithm with RBF-Polynomial (degree 4) kernel for training and testing data in rainfall forecasting. This study focused on monthly rainfall data collected from Mataram City, Indonesia. We developed a hybrid BP-RVM algorithm as the main algorithm that offers a predictive approach to compare the trainlm and trainrp functions. We conducted 20 trials with combinations of learning, momentum, and gamma-RBF at internal values of 0.01-0.9. The training results from trainrp with more than 118 iterations yielded the best performance with learning rate 0.8 and momentum 0.2; MSE value of 2,236.25 and RMSE of 47.29. These results indicate a relatively low error rate for the proposed method. In contrast, the trainlm method, which only requires 18 iterations with a learning rate of 0.6 and momentum of 0.4, produces an MSE of 2,689.25 and RMSE of 51.86, showing its efficiency in reducing the computation time but with a slightly higher error rate than trainrp. Overall, the trainrp method was more accurate in capturing actual rainfall patterns with lower error rates, whereas the trainlm method exhibited good stability but greater sensitivity to parameter variations. This comparative analysis highlights the potential of trainrp to achieve more precise rainfall predictions within the study area