Publication Search

Chronic diseases such as chronic kidney disease (CKD), diabetes, and heart disease remain major causes of mortality worldwide, highlighting the need for accurate and interpretable diagnostic models. However, conventional machine learning methods often face challenges of limited generalization, feature redundancy, and class imbalance in medical datasets. This study proposes an integrated classification framework that unifies three complementary feature paradigms: classical tabular attributes, deep latent features extracted through an unsupervised Long Short-Term Memory (LSTM) encoder, and quantum-inspired features derived from a five-qubit circuit implemented in PennyLane. These heterogeneous features are fused using a feature-wise attention mechanism combined with an AdaBoost classifier to dynamically weight feature contributions and enhance decision boundaries. Experiments were conducted on three benchmark medical datasets—CKD, early-stage diabetes, and heart disease—under both balanced and imbalanced configurations using stratified five-fold cross-validation. All preprocessing and feature extraction steps were carefully isolated within each fold to ensure fair evaluation. The proposed hybrid model consistently outperformed conventional and ensemble baselines, achieving peak accuracies of 99.75% (CKD), 96.73% (diabetes), and 91.40% (heart disease) with corresponding ROC AUCs up to 1.00. Ablation analyses confirmed that attention-based fusion substantially improved both accuracy and recall, particularly under imbalanced conditions, while SMOTE contributed minimally once feature-level optimization was applied. Overall, the attention-guided AdaBoost framework provides a robust and interpretable approach for clinical risk prediction, demonstrating that integrating diverse quantum, deep, and classical representations can significantly enhance feature discriminability and model reliability in structured medical data.

This study examines the application of Educational Data Mining (EDM) to predict the academic per-formance of first-year students at the Catholic University of Bukavu and the Higher Institute of Edu-cation (ISP) in the Democratic Republic of Congo. The primary objective is to develop a model that can identify at-risk students early, providing the university with a tool to enhance student support and academic guidance. To address the challenges posed by data imbalance (where successful cases outnumber failures), the study adopts a hybrid methodological approach. First, the SMOTE algorithm was applied to balance the dataset. Then, a stacking classification model was developed to combine the predictive power of multiple algorithms. The variables used for prediction include the National Exam score (PEx), the secondary school track (Humanities), and the type of prior institution (public, private, or religious-affiliated schools), as well as age and sex. The results demonstrate that this approach is highly effective. The model is not only capable of predicting success or failure but also of forecasting students' performance levels (e.g., honors or distinctions). Moreover, the use of the Apriori association rule mining algorithm allowed the identification of faculty-specific success profiles, transforming prediction into an interpretable decision-support tool. This research makes several significant contributions. Practically, it provides the University of Bukavu with a tool for student orientation and early risk detection. Methodologically, it illustrates the effectiveness of a combined approach to EDM in an African context. However, the study acknowledges certain limitations, including the non-public nature of the data and the geographical specificity of the sample. It therefore proposes avenues for future research, such as the integration of Explainable AI (XAI) techniques for more refined and transparent analysis of the results.

Sleep is a fundamental human need that plays a crucial role in maintaining both physical and mental health. Poor sleep quality can trigger a variety of health problems, ranging from decreased concentration to an increased risk of chronic diseases. The complexity of factors influencing sleep quality—such as stress levels, heart rate, blood pressure, physical activity, and lifestyle—makes its assessment difficult through direct observation alone. Therefore, data mining approaches are increasingly utilized to identify relevant patterns in sleep-related data. This study aims to compare the performance of the C4.5 (Decision Tree) algorithm and the Naïve Bayes algorithm in predicting sleep quality using the Sleep Health and Lifestyle dataset, which contains information from 374 respondents. The research method applied is a quantitative comparative approach employing classification techniques with 10-fold cross-validation to ensure robust evaluation. Model performance is assessed using accuracy, precision, and recall metrics to provide a comprehensive understanding of the effectiveness of each algorithm. The findings indicate that the C4.5 algorithm achieves an accuracy of 96.26% and offers advantages in terms of interpretability through its decision tree visualization, enabling easier understanding of variable relationships. In contrast, the Naïve Bayes algorithm demonstrates superior predictive performance, achieving an accuracy of 98.66% along with consistently high precision and recall across nearly all classes. These results suggest that Naïve Bayes is more effective for predictive tasks involving sleep quality, while C4.5 remains highly valuable when the goal is to interpret variable interactions and decision rules. Overall, this research highlights the potential of data mining techniques in health informatics, particularly in improving the understanding and prediction of sleep quality, which in turn can contribute to better prevention and management of sleep-related health issues.

This research was conducted from June to July 2025 in Binjai City, with the primary focus being analyzing the readiness of the Binjai City Regional Disaster Management Agency (BPBD) to implement a flood early warning system utilizing artificial intelligence (AI). The data collection process was conducted through a literature review, which involved reviewing various theories and previous research results regarding the application of AI and Internet of Things (IoT) technology in the context of disaster mitigation. Based on the results of the study, it was found that the use of technologies such as ultrasonic sensors, microcontrollers, fuzzy logic, and automatic notification systems can provide real-time warnings with a high level of accuracy and a fast response. This system enables early detection of rising river levels through automatic measurements, intelligent data processing, and sending notifications to authorities and affected communities within seconds. By integrating historical data and machine learning-based predictions, this system is also able to depict potential flooding before it occurs, providing a longer response time for evacuation. However, the readiness of the Binjai City BPBD still faces various challenges, such as limited digital infrastructure, the need for human resource training in the technology field, and inadequate budget allocation. Therefore, cross-sector collaboration and ongoing policy support are needed for optimal implementation of this system. The use of AI and IoT in early warning systems is not only technically relevant but also urgent in the face of increasing climate change and flood risks. A strategy involving cross-sector collaboration between government, academia, and the private sector is needed to develop an adaptive and sustainable early warning system.

The issue of determining the number of students' graduation times is one of the important indicators in transmitting the quality and effectiveness of the higher education process in universities. The rate of on-time graduation not only impacts accredited institutions, but also becomes a concern for campus management in designing learning strategies and academic guidance. This study aims to apply and compare two classification algorithms in data mining, namely C4.5 and K-Nearest Neighbor KNN, in predicting the accuracy of students' graduation times. Predictions are made based on academic attributes such as Grade Point Average GPA, number of credits that have been achieved, and Semester Grade Point Average IPS as input variables. The method used in this study is Knowledge Discovery in Database KDD which includes data selection, preprocessing, transformation, data mining, and evaluation of results. The study was conducted using the RapidMiner tool, with a dataset of 279 Informatics Study Program students from the 2015 to 2019 intake. The data was classified into two categories: "graduated on time" and "not graduated on time". The test results showed that the KNN algorithm provided better performance compared to C4.5. KNN produced an accuracy of 76.08%, with a precision of 73.11% and a recall of 41.92%. Meanwhile, the C4.5 algorithm produced an accuracy of 73.49%, with a precision of 64.62% and a recall of 41.89%. This difference in accuracy indicates that KNN is more effective in capturing patterns in the data and providing more accurate predictions in this context. Thus, the KNN algorithm can be considered a more optimal method to assist universities in predicting potential student admissions in a timely manner, thus enabling early intervention for students at risk of late graduation. This research also contributes to the development of data mining-based academic decision support systems in higher education.

The study presents a comprehensive framework for optimizing customer retention budget by integrating clustering, classification, and mathematical optimization techniques. The study begins with the IBM Telco dataset, which is prepared through data cleansing, encoding, and scaling.  In the preliminary phase, customer segmentation is performed using K-Means clustering, with k = 3 and k = 4 identified as optimal based on the elbow method and Silhouette score. The configurations produced three (Premium, Standard, Low) and four (Premium, Standard Plus, Standard, Low) customer segments based on purchase preferences, which served as input features for churn prediction. In the second phase, the dataset was divided into training and test sets in an 80:20 ratio, followed by data balancing using the Synthetic Minority Over-sampling Technique (SMOTE) and Edited Nearest Neighbors (ENN). Multiple classification algorithms were evaluated, including Naive Bayes (NB), Random Forest (RF), Categorical Boosting (CatBoost), Light Gradient Boosting Machine (LightGBM), Extreme Gradient Boosting (XGBoost), Gradient Boosting (GB), Support Vector Machine (SVM), Logistic Regression (LR), K-Nearest Neighbors (KNN), and Multi-Layer Perceptron (MLP) using F1-score as the performance metric. CatBoost and LightGBM, with k values of 3 and 4, respectively, were the highest-performing classification models, with only minimal differences in performance.    Ultimately, customer segmentation established customer prioritization, whereas churn prediction assessed customer churn likelihood. Four distinct configurations were assessed utilizing mixed-integer linear programming (MILP) to optimise retention budget allocation within uniform budget constraints, discount amounts, and churn thresholds. In both the k=3 and k=4 scenarios, CatBoost surpassed LightGBM, with CatBoost at K=3 effectively discounting 66% of at-risk consumers across all three segments, hence improving the intervention's efficacy and budget allocation, making it the ideal choice for maximizing customer retention. The results demonstrate the importance of segmentation in enhancing retention budgeting and budget optimization, particularly concerning parameter sensitivity.

Bitcoin is one of the most prominent digital assets in the modern financial era due to its high volatility and huge profit potential. However, its extreme price volatility also makes it a high-risk asset, so a reliable forecasting approach is needed to help investors make more rational decisions. This study aims to forecast Bitcoin price using the Moving Average (MA) method, specifically MA3, by utilizing monthly historical data of Bitcoin price in USD currency obtained from investing.com website. The MA3 method was chosen for its ability to smooth out short-term fluctuations and identify the direction of price trends. The forecasting process is performed by calculating the average of the last three months' prices for each point in time and compared to the actual price to evaluate its accuracy. The evaluation is done using various prediction error metrics, namely Error, Absolute Error, Squared Error, and Percentage Error. The results of the analysis show that the MA method provides a fairly representative picture of price trends and can be used as an early indicator in short-term investment strategies. Thus, the Moving Average method proves to be a simple but effective prediction tool, especially for novice investors in the dynamic crypto asset market.

Dengue Fever (DF) continues to be a major public health threat in Indonesia, especially in urban areas with high population density, such as Purwokerto City. This study aims to develop a predictive model to identify high-risk areas for DF outbreaks by integrating Machine Learning (ML) algorithms and Geographic Information Systems (GIS). The research utilizes historical dengue case data, meteorological parameters (rainfall, temperature, humidity), and population density as predictive variables. Three ML classification algorithms—Naïve Bayes, Logistic Regression, and Support Vector Machine (SVM)—were implemented to develop risk prediction models. Extensive data preprocessing, feature selection, and spatial integration were applied to ensure model robustness. The results show that the SVM model outperformed other methods, achieving the highest accuracy, precision, recall, and F1-score in classifying dengue risk zones. Risk maps generated through GIS visualization successfully identify priority areas for targeted interventions. The novelty of this research lies in the combination of local epidemiological data, multi-algorithm comparison, and geospatial mapping to improve early warning systems for DF in Purwokerto. This integrated approach is expected to support more effective prevention strategies and enhance public health preparedness.

This study explores the development and application of a digital twin-driven cybersecurity risk assessment model for Industrial Internet of Things (IIoT) networks. The increasing complexity and interconnectivity of IIoT systems have expanded the attack surface, making them vulnerable to a wide range of cyber threats. The digital twin model addresses this challenge by creating real-time virtual replicas of physical systems, which can simulate and predict network vulnerabilities and attack vectors. The model uses machine learning algorithms and real-time data to simulate cyberattacks, including Distributed Denial of Service (DDoS), malware, and data breaches. By providing continuous monitoring and dynamic risk predictions, the digital twin model enhances the resilience of IIoT networks compared to traditional cybersecurity frameworks. The findings indicate that the model's ability to predict potential cyber threats and simulate various attack scenarios provides a more proactive and accurate approach to cybersecurity in IIoT environments. Additionally, the study highlights key mitigation strategies, including adaptive security mechanisms, real-time anomaly detection, and the use of lightweight encryption for resource-constrained devices. Despite its effectiveness, challenges such as computational requirements, integration with legacy systems, and scalability were identified. This research underscores the strategic importance of digital twin models in securing IIoT systems and advancing Manufacturing 4.0 ecosystems. Future research should focus on enhancing model accuracy, expanding its application to diverse industrial sectors, and improving interoperability with legacy systems to further strengthen the security posture of IIoT networks.

This study examines the role of smart technology, particularly Artificial Intelligence (AI) and the Internet of Things (IoT), in strengthening economic resilience in the face of climate change impacts. Using a qualitative descriptive approach with a literature stufy method, secondary data was obtained from scientific journals, books, proceedings, and relevant online articles. The analysis was conducted through reduction, categorization, and thematic analysis of the relevant literature. The results show that AI contributes significantly to improving economic efficiency and risk prediction compabilities. While IoT strengthens connectivity and automation that support supply chain stability, the intregration of AI and IoT in the agricultural sector significantly increases productivity and food security. In addition, smart technology is also an effective mitigation tool against exctreme climate variations that impact the economy and society. This study emphasizes the importance of cross-sector collaboration and digital infrastructure investment to build adaptive and sustainable economic resilience. The implication of this research provide a basis for policy strategies and digital innovation in an era of increasing dynamic climate change.

This study examines the factors that influence financial distress in companies affiliated with Israel, focusing on the roles of profitability, liquidity, leverage, sales growth, and firm size. The research is driven by the phenomenon of boycotts caused by geopolitical conflicts involving Israel, which have impacted the financial performance of several companies, particularly in Indonesia. The study uses a quantitative approach, analyzing a sample of companies listed on the Indonesia Stock Exchange (IDX) that are affiliated with Israel during the 2023-2024 period. The data consists of quarterly financial statements, which are analyzed using the Altman Z-Score bankruptcy prediction model. The findings show that profitability and liquidity have a significant effect on financial distress, while leverage and sales growth have a smaller impact. Firm size is also found to reduce the risk of financial distress. These results suggest that companies linked to Israel are more vulnerable to financial risks due to boycotts triggered by international political tensions.

The importance of data mining in Indonesia is increasing along with the growth of big data in various strategic sectors. Data mining plays an important role in transforming complex data into useful information to support data-driven decision making, which is urgently needed in the face of competitive challenges and operational complexity. This research aims to examine the development of data mining techniques and applications in Indonesia over the last decade (2015-2024). Through a systematic literature review approach, data was collected from academic publications in SCOPUS indexed databases. From the initial 95 papers found, a further selection was made based on accessibility, title, and abstract until 64 papers were included in the article review. The results show that techniques such as K-Means, Naive Bayes, and Decision Tree are most commonly used. In the business sector, clustering through K-Means is widely applied for market segmentation and consumer pattern analysis. The healthcare sector mainly utilizes classification techniques, such as Naive Bayes and Decision Tree, for disease risk prediction and early diagnosis. Meanwhile, the education sector uses data mining to assess student performance and predict potential dropouts, assisting institutions in optimizing learning strategies.

The life insurance industry plays a strategic role in the national financial system, not only as a provider of protection against life risks such as premature death or critical illness, but also as an instrument of long-term fund accumulation. Increased public awareness of the importance of risk protection has driven significant growth in the number of active policies. This condition has a direct impact on the risk exposure of claims that must be carefully managed by insurance companies. One of the main challenges in risk management is to accurately estimate the number of claims in a certain period, to support premium setting, technical reserve planning, and maintain the company's financial stability. This study aims to examine the use of Poisson regression model in estimating the frequency of life insurance claims based on the number of active policies in life insurance company. The data used is simulative and represents an exponential relationship between the number of policies and claims. The model is analyzed using the Maximum Likelihood Estimation (MLE) approach and evaluated through goodness-of-fit indicators such as deviance, Pearson chi-square, log-likelihood, and Mean Squared Error (MSE). The results of the analysis show that the Poisson regression model can capture the significant relationship pattern between the number of active policies and claims, and provide accurate prediction results. Thus, Poisson regression is proven to be a relevant and applicable statistical method in supporting strategic decision-making in insurance companies, especially in the context of data-driven risk management.

. The Framingham Risk Score (FRS) assesses coronary heart disease (CHD) risk and predicts acute coronary events. Metabolic markers like LDL cholesterol, fasting blood glucose, uric acid, triglycerides, and TG/HDL ratio play critical roles in atherosclerosis and cardiovascular risk. Elevated LDL cholesterol, fasting blood glucose, and uric acid contribute to plaque formation, inflammation, and vascular damage, while high triglycerides and low HDL cholesterol exacerbate atherogenesis. This study explores the relationship between these markers and FRS to enhance CHD risk prediction and support targeted cardiovascular interventions. This study analyzed LDL cholesterol, fasting blood glucose, uric acid, triglycerides, and TG/HDL ratio with Framingham Risk Score in 85 participants, excluding those with incomplete data or chronic illnesses. The analysis found significant correlations between metabolic parameters and the 10-year myocardial infarction risk. LDL cholesterol, triglycerides, and uric acid showed moderate positive associations with cardiovascular outcomes, while the triglyceride-to-HDL ratio and fasting blood glucose had weaker but significant correlations. These findings highlight lipid profiles and metabolic markers as key contributors to cardiovascular risk. This study highlights significant correlations between LDL cholesterol, fasting blood glucose, uric acid, triglycerides, and the triglyceride/HDL ratio with 10-year cardiovascular risk. These findings emphasize the importance of lipid profiles, glycemic control, and metabolic markers in predicting coronary outcomes and guiding targeted preventive interventions for improved cardiovascular risk management.

Thyroid cancer is the most common endocrine malignancy, with a steadily increasing incidence rate. Although the overall survival rate is relatively high, the risk of recurrence after definitive treatment such as Radioactive Iodine (RAI) therapy remains a significant clinical challenge. Predicting recurrence risk is crucial for optimizing monitoring strategies and interventions. With advances in technology, machine learning (ML) approaches are increasingly utilized to support medical predictions, including the recurrence of thyroid cancer. This study aims to evaluate the performance of four classification algorithms—Logistic Regression, XGBClassifier, Random Forest Classifier, and Voting Classifier—in predicting thyroid cancer recurrence using the Thyroid Cancer Recurrence After RAI Therapy dataset, which consists of 383 patient records and 13 key clinical attributes. The evaluation was conducted using accuracy, precision, recall, F1-score, and area under the curve (AUC) metrics. The results show that the XGBClassifier is the best-performing model with an accuracy of 97.4% and an AUC of 0.95, demonstrating superior performance in handling the minority class. This research is expected to contribute to the development of more effective machine learning–based clinical decision support systems for predicting thyroid cancer recurrence after therapy.

The increasingly advanced era is certainly followed by various developments in the fields of economy, social, culture and even technology. Technological progress in various countries can be seen from the existence of a technology called AI. Artificial Intelligence or AI is an artificial intelligence technology that has the ability to solve problems like humans. In practice, AI will simulate human intelligence with the ability to recognize images, write and even make predictions based on data. Lately, AI has begun to be misused by individuals who feel they are benefiting. In Indonesia, the misuse of AI technology has been rampant where perpetrators use AI for fraud. An act that uses technology as the main weapon of crime is included in the category of cyber crime. The rampant case of cyber crime in Indonesia is the act of voice imitation or voice cloning. The perpetrators in committing their crimes usually use a technological intermediary that can change their voice so that it is similar to the voice of relatives, friends, family or important people or public figures recognized by the victim. Voice imitation or known as voice cloning is the ability to imitate a voice that is similar to the original voice. Voice cloning using AI technology is an imitation of the human voice with an extraordinary level of accuracy of similarity, both in intonation, tone, and also voice patterns. The law on information and electronic transactions contains various regulations regarding information and electronic transactions, but the law does not explain in detail the misuse of AI in voice cloning because there are still limitations in the ITE law regarding criminal acts of voice cloning fraud and until now criminal acts of voice cloning are still rampant.

In geotechnical engineering, building robust structures is crucial to ensure the bearing capacity of structures against external forces, so making sure soil strength and unreliable build cost and duration prediction are also very important and preliminary aspects of any construction project. Therefore, in this first-of-its-kind modern examine, the capability of various artificially intelligent (AI)-based models toward reliable forecasting and estimation of preliminary construction expenses, duration, and strength at shear is explored. First, background information about the revolutionary artificial intelligence (AI) technique along with its many distinct models ideal for geotechnical and building engineering problems is presented, The use of AI-based models in the literature for the aforementioned construction and maintenance applications is discussed in a number of current works, together with their benefits, drawbacks, and future directions. Several important input elements that significantly affect the preliminary price of construction, construction time, and soil's shear strength estimation are listed and given through analysis. Finally, some obstacles to employing AI-based models for precise forecasts in these applications are discussed, along with elements influencing the problems with cost overruns. Thus, this work can help civil engineers make effective use of artificial intelligence (AI) to solve difficult and risky tasks. It can also be used to Internet of Things (IoT) environments for self-learning applications like smart architectural health-monitoring systems

Given the increasing prevalence of mental health challenges in digital work settings, especially among IT remote workers, early detection mechanisms have become critically important. This study aims to improve the prediction accuracy of mental health conditions among IT remote workers by integrating feature engineering techniques within machine learning models. Five algorithms consisting of Random Forest, Logistic Regression, K-Nearest Neighbors, Decision Tree, and Naive Bayes were evaluated. The Random Forest model achieved the best performance, with 83% accuracy, 83% precision, 100% recall, and a 90% F1-score, followed closely by Logistic Regression with 82% accuracy. Nevertheless, the results demonstrate the feasibility of applying machine learning to support the early detection of mental health risks, offering a strong foundation for future research in predictive analytics and the development of intelligent support systems within digital work environments.

Health risk stratification is crucial for preventive healthcare, yet existing models often rely on binary classification generalized disease prediction, neglecting personalized health indicators and graded risk levels. Many studies apply feature selection techniques like Relief and Univariate Selection without quantifying the weighted impact of features. To address these gaps, this study introduces a Big Data-driven Health Index (HI) framework using PySpark for scalable health risk stratification. The HI is computed as a weighted sum of health-related features using SHAP Analysis, XGBoost, Random Forest, and Correlation Analysis. PySpark enables efficient processing of large-scale health data, and individuals are classified into Low and High Risk. Optimal classification thresholds are determined using the Youden Index from the ROC curve to balance sensitivity and specificity. Personalized health recommendations are generated based on risk categories to guide preventive interventions. Performance evaluation reveals that Correlation Analysis achieves 100% precision and 98.90% recall, outperforming other methods. SHAP prioritizes recall but has low precision, while XGBoost and Random Forest improve precision but struggle with recall. By leveraging Big Data techniques with PySpark, this study enhances computational efficiency, scalability, and classification accuracy, addressing prior research limitations and providing a robust data-driven approach to personalized health monitoring.

Customer churn is a major challenge in e-commerce, directly affecting revenue and profit. This study aims to develop a machine learning model using XGBoost to predict churn probability. To handle class imbalance, SMOTE was applied as a resampling method, and hyperparameter tuning was performed to enhance performance. The model was evaluated using the F2-score, prioritizing recall while maintaining precision. The results show that the XGBoost model with SMOTE achieves strong performance, with an F2-score of 0.849 on the tuned test data. This model can help businesses identify at-risk customers early, enabling proactive retention strategies.