Kinetik: Game Technology, Information System, Computer Network, Computing, Electronics, and Control
Not a member yet
424 research outputs found
Sort by
Classification of Arrhythmia Electrocardiogram Signals Using Kernel Principal Component Analysis and Naive Bayes
No full textArrhythmia is a cardiovascular disorder commonly detected through electrocardiogram (ECG) signal analysis. However, classifying arrhythmias based on ECG signals remains challenging due to signal complexity and individual variability. This study aims to develop a more accurate and efficient method for arrhythmia classification. The proposed method utilizes Kernel Principal Component Analysis (KPCA) and the naïve Bayes algorithm to classify arrhythmic ECG signals. KPCA is chosen for its ability to reduce data dimensionality, facilitating the processing of complex ECG signal and improving classification accuracy by minimizing noise. The naïve Bayes algorithm is chosen for its simplicity and computational speed, as well as its effective performance, even with limited data. ECG signals are processed using KPCA to reduce data dimensionality and extract relevant features. Subsequently, the naïve Bayes algorithm is then applied to classify the ECG signals into four categories: Premature Atrial Contraction (PAC), Premature Ventricular Contraction (PVC), Left Bundle Branch Block (LBBB), and Right Bundle Branch Block (RBBB). The model's performance is evaluated using metrics such as accuracy, sensitivity, specificity, precision, and F1-score. The naïve Bayes model achieves an overall accuracy of 97.67%, with the highest performance observed in the RBBB class at 99.33%. Additionally, the F1-scores across all classes range from 96.62% to 98.57%, demonstrating the model's capability in detecting arrhythmias effectively. These results indicate that the combination of KPCA and naïve Bayes is effective for arrhythmic ECG signals classification
A Metaheuristic Wrapper Approach to Feature Selection with Genetic Algorithm for Enhancing XGBoost Classification in Diabetes Prediction
No full textThis study addressed the problem of selecting the most relevant features for improving the accuracy of diabetes classification using health indicator data. The research focused on a binary classification task based on the Behavioral Risk Factor Surveillance System dataset, which comprised over seventy thousand records and twenty-one predictive features related to individual health behaviors and conditions. A metaheuristic wrapper approach was developed by integrating a Genetic Algorithm for feature selection with an XGBoost classifier to evaluate the predictive quality of each feature subset. The fitness function was defined as the average classification accuracy obtained through cross-validation. In addition to feature selection, hyperparameter optimization of the XGBoost model was carried out using a Bayesian-based search strategy to further enhance performance. The proposed method successfully identified a subset of fourteen optimal features that contributed most significantly to the prediction of diabetes. The final model, combining the selected features and optimized parameters, achieved an accuracy of 0.753, outperforming both the baseline models trained on all features and models using features selected through deterministic methods. These results confirmed the effectiveness of combining evolutionary feature selection with model tuning to build efficient and interpretable predictive models for medical data classification. This approach demonstrated a practical solution for managing high-dimensional data in the context of chronic disease prediction
Development of a Web-Based Information System for Real-Time Fainting Detection Using YOLO in Smart Healthcare
No full textLoss of consciousness (fainting) is a critical condition that requires prompt treatment, especially in the context of elderly health services and independent patient care. This research aims to develop a web-based information system that is able to detect fainting events in real-time using the You Only Look Once (YOLO) algorithm version 11, which is one of the latest approaches in deep learning-based object detection. The system is designed to monitor video from the surveillance camera directly, make visual inferences of the patient's posture, and provide automatic notifications if a loss of consciousness condition is detected. The dataset was obtained from the Roboflow platform and consists of 9,081 annotated images representing the fainting position. The YOLOv11 model was trained and tested using training data sharing, validation, and testing methods. The test results showed that the model achieved mAP, precision, recall and F1-score values of 98.70%, 98.00%, 97.30% and 97.65%, respectively. The developed information system is able to display the detection visually through the bounding box on the dashboard and record the time of the incident. With this performance, this system shows great potential in improving patient safety through intelligent monitoring and automated response in hospital, nursing home, and residential environments. This research also opens up opportunities for the development of more adaptive AI-based health monitoring systems and computer vision in the future
The Evolution of Image Captioning Models: Trends, Techniques, and Future Challenges
No full textThis study provides a comprehensive systematic literature review (SLR) of the evolution of image captioning models from 2017 to 2025, with a particular emphasis on the impending problems, methodological enhancements, and significant architectural developments. The evaluation is guided by the increasing demand for precise and contextually aware image descriptions, and it adheres to the PRISMA methodology. It selects 36 relevant papers from reputable scientific databases. The results indicate a significant transition from traditional CNN-RNN models to Transformer-based architectures, which leads to enhanced semantic coherence and contextual comprehension. Current methodologies, such as prompt engineering and GAN-based augmentation, have further facilitated generalization and diversity, while multimodal fusion solutions, which incorporate attention mechanisms and knowledge integration, have improved caption quality. Additionally, significant areas of concern include data bias, equity in model assessment, and support for low-resource languages. The study underscores the fact that modern vision-language models, such as Flamingo, GIT, and LLaVA, offer robust domain generalization through cross-modal learning and joint embedding. Furthermore, the efficacy of computing in restricted environments is improved by the development of pretraining procedures and lightweight models. This study contributes by identifying future prospects, analyzing technical trade-offs, and delineating research trends, particularly in sectors such as healthcare, construction, and inclusive AI. According to the results, in order to optimize their efficacy in real-world applications, future picture captioning models must prioritize resource efficiency, impartiality, and multilingual capabilities
PID Controller-Based Simulations for Controlling Inverter Voltage to Enhance Power in a Microgrid
Get PDFAn inverter is a device that converts direct current (DC) into alternating current (AC), which is crucial in various applications, including solar power systems, uninterruptible power supplies (UPS), and electric motor control. Accurate and stable voltage control of the inverter is essential to ensure the performance and reliability of the system. The Proportional-Integral-Derivative (PID) control method is one of the most commonly used control techniques due to its simplicity and effectiveness across different control systems. This study focuses on the implementation of inverter voltage control using a PID controller. The PID controller is designed to regulate the inverter's output voltage, ensuring stability even in the presence of disturbances or load variations. In this research, the mathematical model of the inverter and the PID control system is developed and simulated using MATLAB/Simulink software. The simulation results demonstrate that the PID controller effectively maintains the inverter's output voltage, providing a rapid transient response with minimal overshoot. The application of the PID controller to the inverter also shows improvements in system stability and a reduction in steady-state error. Furthermore, precise tuning of the PID parameters is a key factor in achieving optimal control performance. This research makes a significant contribution to the field of inverter control by demonstrating the effectiveness of the PID controller in regulating the inverter's output voltage. The practical implementation of PID controllers on inverters is expected to enhance the efficiency and reliability of power systems that utilize inverters
Analyzing Perceptron Algorithm for Global Gold Price Prediction using Quantum Computing Approach
Get PDFThe price of gold has garnered significant attention in the world of finance and investment due to its role as a safe haven asset and an indicator of global economic stability. An inherent risk of investing in gold is the daily fluctuation in prices, which can rise, fall, or remain stable. Investors are constantly seeking accurate ways to predict gold price movements in order to make informed investment decisions. While classic algorithms like artificial neural networks have been used for gold price prediction, they often struggle with analyzing complex data and identifying the hidden patterns within large datasets. It is widely acknowledged that accurately and consistently predicting the gold price movements, exchange rate, and whether the gold price will rise or fall is very challenging. To address this challenge, this study explored the use of quantum perceptron algorithm for predicting global gold prices. This approach harnesses the principles of quantum computing to improve the efficiency and performance of neural network models. Quantum computers can perform multiple computations simultaneously, enabling the solution of problems that are difficult for classical computers. This study utilized global gold data from January 2018 to December 2022, with 80:20 split of training and testing data; data from January 2018 to December 2021 for training and data from January 2022 to December 2022 for testing. This study aims to offer insights into the potential and application of quantum algorithms in predicting gold prices. The research involved an analysis of global gold price predictions using the quantum perceptron algorithm and quantum computing
Performance Comparison of Machine Learning Algorithms for Ikat Weaving Classification
Get PDFIkat weaving is a rich traditional heritage of Kota Kediri, Indonesia, with a diverse array of intricate motifs that reflect the cultural richness of the region. As new motifs emerge and information about older designs fades, manual identification becomes time-consuming and difficult. This study leverages machine learning technology, specifically XGBoost, Random Forest, and Neural Network algorithms, to automate the classification of these weaving patterns. The dataset consisted of 600 images, split into 480 images (80%) for training and 120 images (20%) for testing, representing four distinct weaving motifs: "Gumul Weaving, Bolleches Weaving, Kuda Kepang Weaving, and Sekar Jagad Weaving." The study achieves high accuracy, with precision, recall, and F1-score all reaching 100%, underscoring its potential to not only improve the efficiency of motif identification, but also play a crucial role in preserving and promoting Indonesia's cultural heritage. Future research should focus on further optimizing these algorithms and expanding datasets to capture a broader range of ikat motifs. Additionally, enhancing the application of this model can contribute to a deeper understanding and broader appreciation of Kota Kediri’s cultural wealth through digital platforms
Detecting Acute Lymphoblastic Leukemia in Blood Smear Images using CNN and SVM
Get PDFAcute Lymphoblastic Leukemia (ALL) is a common and aggressive subtype of leukemia that predominantly affects children. Accurate and timely diagnosis of ALL is critical for successful treatment, but it is hindered by the limitations of manual examination of peripheral blood smear images, which are prone to human error and inefficiency. This study proposes an improved diagnostic approach by integrating the EfficientNet architecture with a Support Vector Machine (SVM) classifier to enhance classification accuracy and address the performance inconsistencies of standalone EfficientNet models. Additionally, a novel CNN-based model with a reduced number of parameters is developed and evaluated. A dataset comprising 3.256 peripheral blood smear images across four classes (benign, early, pre and pro) was used for training and testing. The EfficientNet-SVM models achieved a peak accuracy of 97.35% using the EfficientNet-B3 architecture, surpassing previous studies. The improved CNN model achieved the highest accuracy of 99.18% while reducing parameters by 59.5% compared to the best prior models, with a negligible accuracy decrease of only 0.67%. These findings highlight the potential of combining EfficientNet with SVM and the efficiency of the improved CNN model for automated ALL detection, paving the way for more reliable, cost-effective, and scalable diagnostic tools
Spoon Stabilization for Essential Tremor Patients Using PID Control Optimized by PSO
No full textEssential tremor is a neurological disorder that causes uncontrollable hand tremors, interfering with daily activities such as eating. This study aims to develop a spoon stabilization system controlled by a Proportional-Integral-Derivative (PID) controller, which was tuned using Particle Swarm Optimization (PSO) and the Cohen-Coon method for performance comparison. The system utilized an inertial measurement unit to detect tremors, while a Kalman filter was used to reduce noise before a microcontroller controlled a servo motor to stabilize the spoon. The system was evaluated through simulations and hardware implementation, and its performance was assessed based on rise time, overshoot, delay time, and settling time. The results showed that the Kalman filter significantly reduced noise, lowering the average pitch angle error deviation from 1.028° to 0.037° and the roll angle error from 0.822° to 0.031°. The PSO-based tuning outperformed the Cohen-Coon method in response speed and system stability, achieving a faster rise time (0.09 s for roll, 0.34 s for pitch), a shorter settling time (0.74 s for roll, 0.59 s for pitch), and a lower delay time (0.1 s for roll, 0.15 s for pitch). However, the Cohen-Coon method resulted in a lower overshoot for the roll angle (6.08%) compared to the PSO-based tuning (11.98%). The findings suggest that implementing a PID controller optimized via PSO is a viable approach for spoon stabilization in individuals with essential tremor
Exploiting Vulnerabilities of Machine Learning Models on Medical Text via Generative Adversarial Attacks
No full textSignificant developments in artificial intelligence (AI) technology have fueled its adoption across a range of fields. The use of AI, particularly machine learning (ML), has expanded significantly in the medical field due to its high diagnostic precision. However, the AI model faces a serious challenge to handle the adversarial attacks. These attacks use perturbed data (modified data), which is unnoticeable to humans but can significantly alter prediction results. This paper uses a medical text dataset containing descriptions of patients with lung diseases classified into eight categories. This paper aims to implement the TextFooler technique to deceive predictive models on medical text against adversarial attacks. The experiment reveals that three ML models developed using popular approaches, i.e., transformer-based model based on Bidirectional Encoder Representations from Transformers (BERT), Stack Classifier that combines three traditional machine learning models, and individual traditional algorithms achieved the same classification accuracy of 99.98%. The experiment reveals that BERT is the weakest model, with an attack success rate of 76.8%, followed by traditional machine learning methods and the stack classifier, with success rates of 28.73% and 5.21%, respectively. This implies that although BERT classification demonstrates good performance, it is highly vulnerable to adversarial attacks. Therefore, there is an urgency to develop predictive models that are robust and secure against potential attacks