Environmental and Economic Clustering of Indonesian Provinces: Insights from K-Means Analysis

Leuser Journal of Environmental Studies, Год журнала: 2024, Номер 2(1), С. 41 - 51

Опубликована: Апрель 29, 2024

Indonesia's archipelago presents a distinctive opportunity for targeted sustainable development due to its complex interplay of economic advancement and environmental challenges. To better understand this dynamic identify potential areas focused intervention, study applied K-means clustering 2022 data on the Air Quality Index (AQI), electricity consumption, Gross Regional Domestic Product (GRDP). The analysis aimed delineate provinces into three distinct clusters, providing clearer picture varying levels impact across nation's diverse islands. Each cluster reflects specific dynamics, suggesting tailored policy interventions. results show that in Cluster 1, which exhibit moderate quality lower activity, introduction agricultural enhancements, eco-tourism, renewable energy initiatives is recommended. 2, marked by higher outputs conditions, would benefit from implementation smart urban planning, stricter controls, adoption clean technologies. Finally, 3, includes highly urbanized with robust growth, requires expanded green infrastructure, improved practices, enhanced public transportation systems. These recommendations aim foster balanced growth while preserving integrity Indonesia’s landscapes.

Язык: Английский

---

Enhancing Environmental Quality: Investigating the Impact of Hydropower Energy Consumption on CO2 Emissions in Indonesia

Ekonomikalia Journal of Economics, Год журнала: 2024, Номер 2(1), С. 53 - 65

Опубликована: Апрель 28, 2024

Achieving sustainable environmental quality has become a critical global issue, necessitating the reduction of carbon dioxide (CO2) emissions and greenhouse gas (GHG) to mitigate pollution. Hydropower energy potential play significant role in this effort by providing clean, renewable source that can help reduce reliance on fossil fuels decrease CO2 emissions. This study examines dynamic impact hydropower consumption, economic growth, capital, labor Indonesia's from 1990 2020. Applying Autoregressive Distributed Lag (ARDL) method, findings demonstrate consumption negative effect both short long term, indicating increasing leads Conversely, exhibits positive influence suggesting rise contributes higher levels Indonesia. Furthermore, Granger causality analysis reveals bidirectional relationship between consumption. The robustness ARDL results is confirmed through additional tests using Fully-Modified Ordinary Least Squares (FMOLS), Dynamic (DOLS), Canonical Cointegrating Regressions (CCR) methods. underscore importance promoting for effective management Policymakers should prioritize investments infrastructure, encourage adoption energy-efficient technologies, develop skilled workforce increased force participation.

Язык: Английский

---

A Model-Agnostic Interpretability Approach to Predicting Customer Churn in the Telecommunications Industry

Infolitika Journal of Data Science, Год журнала: 2024, Номер 2(1), С. 34 - 44

Опубликована: Май 27, 2024

Customer churn is critical for businesses across various industries, especially in the telecommunications sector, where high rates can significantly impact revenue and growth. Understanding factors leading to customer essential developing effective retention strategies. Despite predictive power of machine learning models, there a growing demand model interpretability ensure trust transparency decision-making processes. This study addresses this gap by applying advanced specifically Naïve Bayes, Random Forest, AdaBoost, XGBoost, LightGBM, predict dataset. We enhanced using SHapley Additive exPlanations (SHAP), which provides insights into feature contributions predictions. Here, we show that LightGBM achieved highest performance among with an accuracy 80.70%, precision 84.35%, recall 90.54%, F1-score 87.34%. SHAP analysis revealed features such as tenure, contract type, monthly charges are significant predictors churn. These results indicate combining analytics methods provide telecom companies actionable tailor strategies effectively. The highlights importance understanding behavior through transparent accurate paving way improved satisfaction loyalty. Future research should focus on validating these findings real-world data, exploring more sophisticated incorporating temporal dynamics enhance prediction models' applicability.

Язык: Английский

---

Business Confidence in Indonesia: Which Macroeconomic Factors Have Long-Term Impact?

Indatu Journal of Management and Accounting, Год журнала: 2024, Номер 2(1), С. 40 - 54

Опубликована: Июнь 19, 2024

Business confidence refers to the level of optimism or pessimism that business owners have about prospects their companies and overall economy. Thus, focus this study is examine long-term impact various macroeconomic factors—economic growth, government expenditure, interest rates, inflation, exchange composite stock price index—on index in Indonesia by utilizing monthly data from January 2009 December 2022. We employ Dynamic Ordinary Least Squares (DOLS) Fully-Modified (FMOLS) as main methods, with Canonical Cointegrating Regressions (CCR) a robustness check method. The also utilizes pairwise Granger causality tests for comprehensive analysis. findings indicate all factors significantly long term across methodologies. Specifically, economic exert positive impact, while rates negative on index. This evidence emphasizes importance businesses diligently monitor trends understand patterns these indicators so can better anticipate changes sentiment. Taking perspective when making strategic decisions investments advisable, recognizing influence may be more pronounced over time.

Язык: Английский

---

Application of Ensemble Machine Learning Methods for QSAR Classification of Leukotriene A4 Hydrolase Inhibitors in Drug Discovery

Malacca Pharmaceutics, Год журнала: 2024, Номер 2(2), С. 68 - 78

Опубликована: Сен. 20, 2024

Inflammatory diseases such as asthma, rheumatoid arthritis, and cardiovascular conditions are driven by overproduction of leukotriene B4 (LTB4), a potent inflammatory mediator. Leukotriene A4 hydrolase (LTA4H) plays critical role in converting into LTB4, making it prime target for drug discovery. Despite ongoing efforts, developing effective LTA4H inhibitors has been challenging due to the complex binding properties enzyme structural diversity potential inhibitors. Traditional discovery methods, like high-throughput screening (HTS), often time-consuming inefficient, prompting need more advanced approaches. Quantitative Structure-Activity Relationship (QSAR) modeling, enhanced ensemble machine learning techniques, provides promising solution enabling accurate prediction compound bioactivity based on molecular descriptors. In this study, six methods—AdaBoost, Extra Trees, Gradient Boosting, LightGBM, Random Forest, XGBoost—were employed classify The dataset, comprising 636 compounds labeled active or inactive pIC50 values, was processed extract 450 descriptors after feature engineering. results show that LightGBM model achieved highest classification accuracy (83.59%) Area Under Curve (AUC) value (0.901), outperforming other models. XGBoost Forest also demonstrated strong performance, with AUC values 0.890 0.895, respectively. high sensitivity (95.24%) highlights its ability accurately identify compounds, though exhibited slightly lower specificity (61.36%), indicating higher false-positive rate. These findings suggest models, particularly highly predicting bioactivity, offering valuable tools early-stage indicate methods significantly enhance QSAR accuracy, them viable identifying inhibitors, potentially accelerating development anti-inflammatory therapies.

Язык: Английский

---

An Interpretable Machine Learning Strategy for Antimalarial Drug Discovery with LightGBM and SHAP

Journal of Future Artificial Intelligence and Technologies, Год журнала: 2024, Номер 1(2), С. 84 - 95

Опубликована: Авг. 7, 2024

Malaria continues to pose a significant global health threat, and the emergence of drug-resistant malaria exacerbates challenge, underscoring urgent need for new antimalarial drugs. While several machine learning algorithms have been applied quantitative structure-activity relationship (QSAR) modeling compounds, there remains more interpretable models that can provide insights into underlying mechanisms drug action, facilitating rational design compounds. This study develops QSAR model using Light Gradient Boosting Machine (LightGBM). The is integrated with SHapley Additive exPlanations (SHAP) enhance interpretability. LightGBM demonstrated superior performance in predicting activity, an ac-curacy 86%, precision 85%, sensitivity 81%, specificity 89%, F1-score 83%. SHAP analysis identified key molecular descriptors such as maxdO GATS2m contributors activity. integration not only enhances predictive but also provides valuable importance features, aiding approach bridges gap between accuracy interpretability, offering robust framework efficient effective discovery against strains.

Язык: Английский

---

Machine Learning for Early Detection of Dropout Risks and Academic Excellence: A Stacked Classifier Approach

Journal of Educational Management and Learning, Год журнала: 2024, Номер 2(1), С. 28 - 34

Опубликована: Май 24, 2024

Education is important for societal advancement and individual empowerment, providing opportunities, developing essential skills, breaking cycles of poverty. Nonetheless, the path to educational success marred by challenges such as achieving academic excellence preventing student dropouts. Early identification students at risk dropping out or those likely excel academically can significantly enhance outcomes through tailored interventions. Traditional methods often fall short in precision foresight effective early detection. While previous studies have utilized machine learning predict performance, potential more sophisticated ensemble methods, stacked classifiers, remains largely untapped contexts. This study develops a classifier integrating predictive strengths LightGBM, Random Forest, logistic regression. The model achieved an accuracy 80.23%, with precision, recall, F1-score 79.09%, 79.20%, respectively, surpassing performance models tested. These results underscore classifier's enhanced capability transformative settings. By accurately identifying achieve early, institutions better allocate resources design targeted approach optimizes supports informed policymaking, fostering environments conducive success.

Язык: Английский

---

Interpretable Machine Learning for Chronic Kidney Disease Diagnosis: A Gaussian Processes Approach

Indonesian Journal of Case Reports, Год журнала: 2024, Номер 2(1), С. 24 - 32

Опубликована: Июнь 29, 2024

Chronic Kidney Disease (CKD) is a global health issue impacting over 800 million people, characterized by gradual loss of kidney function leading to severe complications. Traditional diagnostic methods, relying on laboratory tests and clinical assessments, have limitations in sensitivity are prone human error, particularly the early stages CKD. Recent advances machine learning (ML) offer promising tools for disease diagnosis, but lack interpretability often hinders their adoption practice. Gaussian Processes (GP) provide flexible ML model capable delivering predictions uncertainty estimates, essential high-stakes medical applications. However, integration GP with interpretable methods remains underexplored. We developed an CKD classification address this knowledge gap combining Shapley Additive Explanations (SHAP). assessed model's performance using three kernels (Radial Basis Function, Matern, Rational Quadratic). The results show that Quadratic kernel outperforms other kernels, achieving accuracy 98.75%, precision 100%, 97.87%, specificity F1-score 98.51%. SHAP values indicate haemoglobin specific gravity most influential features. demonstrate enhances predictive provides robust estimates explanations. This combination supports clinicians making informed decisions improving patient management outcomes Our study connects advanced techniques practical applications, more effective reliable ML-driven healthcare solutions.

Язык: Английский

---

Explainable Deep Learning Approach for Mpox Skin Lesion Detection with Grad-CAM

Heca Journal of Applied Sciences, Год журнала: 2024, Номер 2(2), С. 54 - 63

Опубликована: Сен. 19, 2024

Mpox is a viral zoonotic disease that presents with skin lesions similar to other conditions like chickenpox, measles, and hand-foot-mouth disease, making accurate diagnosis challenging. Early precise detection of mpox critical for effective treatment outbreak control, particularly in resource-limited settings where traditional diagnostic methods are often unavailable. While deep learning models have been applied successfully medical imaging, their use remains underexplored. To address this gap, we developed learning-based approach using the ResNet50v2 model classify alongside five conditions. We also incorporated Grad-CAM (Gradient-weighted Class Activation Mapping) enhance interpretability. The results show achieved an accuracy 99.33%, precision 99.34%, sensitivity F1-score 99.32% on dataset 1,594 images. visualizations confirmed focused relevant lesion areas its predictions. performed exceptionally well overall, it struggled misclassifications between visually diseases, such as chickenpox mpox. These demonstrate AI-based tools can provide reliable, interpretable support clinicians, limited access specialized diagnostics. However, future work should focus expanding datasets improving model's capacity distinguish

Язык: Английский

---

An Explainable Machine Learning Study of Behavioral and Psychological Determinants of Depression in the Academic Environment

Journal of Educational Management and Learning, Год журнала: 2025, Номер 3(1), С. 22 - 31

Опубликована: Май 25, 2025

Depression is a significant and growing concern within academic environments, affecting both students staff due to factors such as pressure, financial stress, lifestyle challenges. This study explores the use of machine learning, specifically Random Forest classifier, predict depression risk among using behavioral, psychological, demographic data. A dataset 27,788 student records was analyzed after thorough preprocessing exploratory data analysis. The model achieved strong performance, with an accuracy 83.52% AUC 0.91, indicating reliable classification status. Local Interpretable Model-agnostic Explanations (LIME) were employed enhance interpretability, revealing key predictive features suicidal ideation, sleep duration, dietary habits. These interpretable insights align existing psychological research provide actionable information for mental health professionals. findings highlight value explainable AI in educational settings, offering scalable transparent approach early detection intervention. Future work should focus on longitudinal integration, multimodal inputs, real-world implementation strengthen model’s utility impact.

Язык: Английский

---