迪克雷大学

Accurate long-term forecasting of greenhouse gas emissions is essential for climate mitigation planning, yet many existing studies either focus on a single gas or prioritize predictive accuracy without sufficient interpretability. To address this gap, this study develops an interpretable machine learning (ML) framework for the joint modeling of carbon dioxide (CO2) and nitrous oxide (N2O) emissions using country-level socio-economic indicators. Five regression-based algorithms, namely Decision Tree (DT), K-Nearest Neighbors (KNN), AdaBoost, Support Vector Machine (SVM), and Random Forest (RF), were benchmarked, and the best-performing model was further optimized using Random Search. The optimized Random Forest achieved the strongest predictive performance for both gases, reaching R2 values of 0.989 for CO2 and 0.982 for N2O, with corresponding Root Mean Square Error (RMSE) values of 40.09 and 3.24, respectively. Cross-validation results also confirmed strong stability, with mean R2 values of 0.987 +/- 0.004 for CO2 and 0.979 +/- 0.006 for N2O. Interpretability analyses based on SHapley Additive exPlanations (SHAP), Local Interpretable Model-agnostic Explanations (LIME), and ablation testing showed that Gross Domestic Product (GDP) is the dominant driver of CO2 emissions, whereas population has a relatively stronger contribution in N2O prediction. In methodological terms, the study contributes a transparent multi-gas framework that integrates model optimization, robustness assessment, and dual-layer explainability. Substantively, the results indicate that CO2 mitigation is more tightly associated with economic decoupling and energy transition, while N2O mitigation requires greater emphasis on agriculture-linked demographic pressures and nitrogen management. The framework is suitable for conditional forecasting and scenario-based policy analysis when future socio-economic pathways are supplied exogenously.

Silene species have various pharmacological properties, including antimicrobial, antioxidant, anti-inflammatory, antitumoral, adaptogenic, immunomodulatory, hepatoprotective, and cytotoxic activities. This study assessed the antibacterial, antioxidant, and cytotoxic activities and the chemical profile of crude methanol extract and sub-extracts prepared from the aerial parts of Silene latifolia for the first time. After a 24-h incubation, the methanol extract exhibited the highest activity against MCF-7 cancer cell line (IC50: 19.25 ± 0.05 µg/mL), whereas the water sub-extract showed the greatest activity against MDA-MB-231 cancer cell line (IC50: 15.20 ± 0.47 µg/mL). After a 48-h incubation, the extracts having the strongest cytotoxic effects against MCF-7 and MDA-MB-231 cells were the water sub-extract (IC50: 14.52 ± 0.86 µg/mL) and the methanol crude extract (IC50: 12.08 ± 0.78 µg/mL), respectively. The methanol extract exhibited the most potent radical-scavenging activity, enhancing total antioxidant status and reducing total oxidant status. The extracts had no antibacterial or antifungal activity against the tested microorganisms. GC–MS analysis revealed that oleic acid, elaidic acid, palmitic acid, and their derivatives were abundant in sub-extracts. The LC–MS/MS analysis demonstrated that the water sub-extract included fumaric acid, quinic acid, aconitic acid, acacetin, and luteolin. This study highlighted that S. latifolia exhibited high antioxidant activity, and that it had a potent cytotoxic effect against breast cancer cells with a high selectivity index.

This study investigates the antioxidant, enzyme inhibitory, and antimicrobial activities of water (WEHL) and ethanol (EEHL) extracts of hop (Humulus lupulus) cones. Phytochemical analyses revealed higher total phenolic content in EEHL (271.52 +/- 0.13 mg GAE/g) than in WEHL (251.84 +/- 0.06 mg GAE/g), as well as higher total flavonoid content (182.56 +/- 0.45 mg QE/g for EEHL versus 179.39 +/- 0.46 mg QE/g for WEHL). Antioxidant activity, determined by DPPH and ABTS assays, showed that EEHL had stronger radical scavenging capacity with IC50 values of 19.13 +/- 4.66 mu g/mL (DPPH) and 12.66 +/- 1.94 mu g/mL (ABTS), compared to WEHL (DPPH: 20.90 +/- 2.39 mu g/mL; ABTS: 32.41 +/- 4.29 mu g/mL). In reducing assays, EEHL also showed better absorbance values in FRAP (0.77 +/- 0.01), CUPRAC (2.09 +/- 0.05), and Fe3+ reducing (1.95 +/- 0.01) tests. EEHL likely outperformed WEHL due to solvent polarity and extraction efficiency. Moderately polar ethanol extracts a broader range of phenolics and flavonoids, including fewer polar bioactive compounds that contribute to antioxidant capacity and enzyme inhibition. This matches higher TPC/TFC in EEHL and explains stronger radical scavenging, reducing power, and multi-enzyme inhibition. Enzyme inhibition studies revealed that EEHL inhibited acetylcholinesterase (IC50: 26.06 mu g/mL), butyrylcholinesterase (IC50: 44.00 mu g/mL), alpha-glycosidase (IC50: 119.31 mu g/mL), and carbonic anhydrase isoenzymes hCA I (IC50: 59.78 mu g/mL) and hCA II (IC50: 21.19 mu g/mL). LC-MS/MS analysis identified major phenolic compounds such as isoquercitrin (3.14 ng/mL), rutin (0.60 ng/mL), and hesperidin (0.43 ng/mL) in EEHL. Antimicrobial screening showed selective activity against Staphylococcus aureus with an inhibition zone of 18.50 +/- 0.58 mm, while no inhibition was observed against Escherichia coli and Candida albicans. These findings provide a solvent-dependent in vitro profile that can guide extraction strategies, support antioxidant and multi-enzyme screening (including hCA I and II), and identify candidates for selective antimicrobial evaluation and further preclinical investigation. Despite extensive use of hop extracts, comparative solvent-dependent profiling that links LC-MS/MS phenolic composition with a broad multi-enzyme inhibition panel, including the less frequently evaluated hCA I/II isoenzymes, remains limited. Therefore, the objective of this study was to systematically compare WEHL and EEHL in terms of phytochemical content and in vitro antioxidant, enzyme inhibitory, and antimicrobial activities. Overall, these results provide a solvent-dependent, comparative in vitro profile of WEHL vs. EEHL that can support antioxidant, multi-enzyme screening (including hCA I and II), and selective antimicrobial assays.

Accurate power prediction and fault detection in photovoltaic (PV) systems are essential for improving energy efficiency and enabling predictive maintenance. This study proposes a novel hybrid regression model based on a stacking ensemble architecture, which integrates multiple machine learning algorithms: histogram-based gradient boosting (HGB), k-nearest neighbors (k-NN), decision tree (DT), random forest (RF), and LightGBM as base learners and employs Ridge regression as the meta-learner. The model was designed to detect complex fault conditions such as partial shading and module-level failures using SCADA-type input features. The performance of the proposed model was evaluated using standard regression metrics (R-2,R- RMSE, MAE), achieving superior results with an R(2 )of 0.9939, RMSE of 12.0184, and MAE of 8.0544. Paired t-tests confirmed the statistical significance of performance improvements over baseline models (p < 0.05). To ensure transparency, explainability analyses were conducted using SHapley Additive exPlanations (SHAP) and local interpretable model-agnostic explanations (LIME), which revealed that fault-related features had the greatest influence on model predictions. Comparative evaluation with recent state-of-the-art approaches demonstrated that the proposed hybrid model is scalable, computationally efficient, and robust under varying environmental and operational conditions. The findings suggest that the model can serve as a reliable and interpretable solution for real-time power forecasting and fault detection in PV systems.

In this study, hat-stiffened composite panels, which are widely employed in the aerospace industry, were fabricated using the vacuum infusion method. To enhance the strength of the hat-stringer composites, a composite I-beam was integrated into the panel, and its impact performance under low-velocity loading was investigated. It was determined that the incorporation of the I-beam into the hat-stiffened composite panels increased the maximum impact force by up to 2.5 times. For the numerical analyses, the LS-DYNA finite element software package was utilized, and the impact test results were examined in three dimensions based on the Hashin failure criterion. Geometric variations that occurred in the hat-stringer and I-beam during the manufacturing process were incorporated into the numerical model. The correlation between the experimental data and the modified numerical results is presented in detail.