High R2 Values Research Articles

Science and Innovation Management: Sustainable Development Goals 8 and 9 as Challenges for Poland

This manuscript continues the series of research publications on Sustainable Development Goals 8 and 9 (SDGs 8 and 9). Innovations play an essential role in the global and national economies, as they contribute to the growth of economic productivity, the creation of new products, services, and jobs, and improve the quality of life. In the context of innovation, economy and artificial intelligence (AI) are the ideological and technological components of the innovation process. The authors hypothesized that the Polish system for assessing the importance of scientific economic journals is ineffective for social development and for creating an innovation-based economy. The research methods are a scientometric analysis of 795,070 scientific sources by the keywords “Economy” and “Artificial Intelligence”, a bibliometric analysis of 219,739 sources by the keywords “Economy”, “Economic productivity”, and “Innovation”, correlation analysis, and testing statistical hypotheses. The study led to the adoption of an alternative statistical hypothesis, which is a strong confirmation of the basic hypothesis. A new model for assessing the weight of scientific journals was created based on the existing system. Two management recommendations were formulated in the rules for determining the weight of journals. The high R2 value indicates that the new model effectively predicts the result. The correlation coefficient of the weight of journals in the new model is more than 98%. When verifying statistical hypotheses, a high significance level of 99.0 was used. The study’s practical significance is in the movement of Polish science to a leading place in the world market of scientific products.

Predictive modeling of CO2 capture efficiency using piperazine solutions: a comparative study of white-box algorithms

The urgency to mitigate carbon dioxide (CO2) emissions and combat climate change has spurred the development of effective CO2 capture technologies. One of the industry’s most well-known CO2 collection processes is CO2 absorption utilizing amine solvents. However, designing a successful amine scrubbing system in power plants requires precise prediction of CO2 absorption in aqueous amine solutions under various operating circumstances. Using aqueous piperazine (PZ) solutions for chemical absorption is promising due to the favorable reactivity of PZ with CO2. In this study, a comprehensive evaluation of PZ solution performance in CO2 capturing, employing the white-box algorithms, namely, Genetic Programming (GP), Gene Expression Programming (GEP), and Group Method of Data Handling (GMDH), was performed. Through extensive experimentation and data analysis, several correlations were developed with high and acceptable R2 values, such as 0.933 for GP, 0.949 for GEP, and 0.889 for GMDH, which shows high accuracy and reliability in predicting the CO2 capture efficiency of PZ solutions under varying operating conditions. The results of sensitivity analysis revealed that CO2 partial pressure increased CO2 absorption, while PZ concentration and temperature had negative and decreasing effects. These insights provide essential guidance for optimizing process conditions to enhance the CO2 capture efficiency. Finally, the leverage method was used to assess the reliability of both experimental and predicted data from white-box algorithms. This analysis identified potential outliers and validated the accuracy of the model’s predictions, enhancing the credibility of developed correlations and demonstrating the robustness of the approach.

Bio-sorption of methylene blue using Datura stramonium leaves as adsorbent

Present study was accomplished to prospect the viability of using the Datura stramonium leaves powder (DS) as an adsorbent to remove the methylene blue from aqueous solution. The physico-chemical characteristics of the studied adsorbent were examined. The optimum parameters such as contact time, particle size, absorbent dose, initial methylene blue concentration, and pH were investigated by performing batch experiments models. The kinetics and the isotherms adsorption were evaluated by varying the initial concentration and using the optimum parameters. The optimum of contact time is 30min, with a removal capacity of 89.60 %. The optimal adsorbent concentration to reach the maximum removal of methylene blue (89.54 %) is 18 g/L. An initial methylene blue concentration of 50 ppm is ideal to reach the maximum capacity of removal (92.72 %). The optimum particle size is 80 mm. The kinetics of the adsorption process are in accordance with the pseudo-second order model. Experimental values of the adsorption capacity are close proximity to the optimum values predicted by the pseudo-second order model. Langmuir, Freundlich, Temkin, Dubinin-Radushkevich, Harkin-Jura and Hasley isotherms were applied to represent the data obtained from the adsorption studies. The highest R2 values were related to Freundlich, Dubinin-Radushkevich and Hasley isotherm models.

Production of biodiesel by using CaO nano-catalyst synthesis from mango leaves extraction

Development and population expansion have the lion's share of driving up the fuel cost. Biodiesel has considerable attention as a renewable, ecologically friendly and alternative fuel source. In this study, CaO nanocatalyst is produced from mango leaves as a catalysis for the transesterification of waste cooking oil (WCO) to biodiesel. The mango tree is a perennial plant, and its fruit holds significant economic worth due to its abundance of vitamins and minerals. This plant has a wide geographical range and its leaves can be utilized without any negative impact on its growth and yield. An analysis was conducted to determine the calcium content in the fallen leaves, revealing a significant quantity of calcium that holds potential for utilization. The catalyst was characterized by different analytic techniques such as XRD, SEM-EDS, FT-IR, and BET analyses. Several parameters impacted on the transesterification process were exploited by conventional transesterification (batch). The result revealed that the optimum reaction was reached at a methanol to oil ratio of 50% w/w, catalyst loading of 3%, temperature of 65℃ and reaction time of 1.5 h with a yield of 93.21%, and the activation energy of the transesterification reaction was found to be 38.906 KJ mol-1. The reaction was verified to be irreversible pseudo-first order based on a linear Arrhenius plot and a high R2 value. The catalyst shows good stability and catalytic activity when it is reused and the yield was found to be 80.293% in the 5th cycle.

Prediction of eco sustainability component using fuzzy Z numbers based ratio analysis and interval type 3 fuzzy logic system

Amid rapid global urbanization, understanding eco-sustainability challenges in megacities like Delhi is crucial both locally and globally. This study introduces a novel multi-criteria group decision-making approach, the Z-Fuzzy Step-Wise Weight Assessment Ratio Analysis (SWARA), to address these complex challenges. This method prioritizes key eco-sustainability components like air quality index (AQI), water quality index (WQI), Temperature, and Precipitation, based on their impact, aiming to enhance resource allocation and environmental quality. Criteria weights are determined using fuzzy analytical hierarchical processing with Z-numbers, which manage decision-makers' uncertainty through a reliability function, illustrating their confidence levels and supporting the adaptability and effectiveness of eco-sustainability measures. A case study is conducted on seven sub-cities in Delhi examining their unique environmental impacts and urban sustainability challenges. Moreover, the most effective component of eco-sustainability say AQI is also dependent on some factors like PM10, PM2.5, SO2, O3, NO2, and CO. Considering these factors as inputs, an interval type-3 fuzzy logic system (IT3FLS) has been constructed to predict AQI, supporting timely interventions and improvements in eco-sustainability management. The IT3FLS model demonstrated exceptional precision across all Delhi sub-cities, with low root means square error (<0.05), and high R2 values (>0.94), affirming its effectiveness. To validate the Z-Fuzzy SWARA rankings, we compared them with existing Z-Fuzzy methods, consistently ranking AQI first. The IT3FLS's AQI predictions were also compared with those from fuzzy logic and advanced machine learning algorithms, demonstrating superior performance. Finally, the study underscores the managerial implications, emphasizing their role in guiding decision-making and strategy implementation.

Modeling Nitrogen Recovery and Water Transport in Gas-Permeable Membranes

This study presents a new modeling approach for nitrogen recovery for gas-permeable membrane (GPM) contactors, including both ammonia and water transport dynamics. A distinct feature of the model is its capacity to model water transport across the membrane, which has been overlooked in most publications. Osmotic pressure differences are used to predict the behavior of ammonia and water transport in the GPM. Experiments carried out to develop, test and calibrate the model examined the dynamics of ammonia and water transport through the GPM at various nitrogen concentrations. Specifically, the GPM contactor was tested for nitrogen recovery from high-strength synthetic wastewaters (2.4-10.6 g N/L) at 35°C and at pH 9. The initial volume of the trapping solution (diluted H2SO4) was 10 times lower than that of the synthetic wastewater, aiming to concentrate the recovered nitrogen. The estimated ammonia transport constant (Km) ranged between (1.2 – 2.1)·10-6 m/s and water transport constant Kw between (2.8 – 8.2)·10-10 m/(s bar). Numerical determination of the model parameters revealed high R² values, demonstrating strong agreement with experimental data.

Synthesis and enhanced adsorption for F− by three novel high entropy LDHs using TEA-assisted hydrothermal technology

Nowadays the design, preparation and application of high entropy LDHs (HE-LDHs) has already become research hotspots. In this paper, three types of high entropy ZnCoCrMgZr-LDHs (HE-LDHs-1, 2 and 3) were successfully prepared for the first time using TEA as alkali source by simple hydrothermal synthesis, and characterized by XRD, TG, SEM, X-ray photoelectron spectroscopy, BET and zeta potential. The effect of triethanolamine (TEA) on the formation of high entropy LDHs and their adsorption behaviour for F− were investigated. Research showed that high content of TEA can maintain and provide more hydroxyl groups required for the formation of LDHs, while the introduction of high valence Zr4+ breaks the regular arrangement of LDHs themselves. Three types of high entropy LDHs have excellent adsorption performance for F− with the maximum adsorption capacities of 113.78 mg/g, 142.12 mg/g and 128.24 mg/g, respectively. They also have strong resistance to anionic interference. The adsorption process is the transition from physical adsorption to chemical adsorption in the F− adsorption process from HE-LDHs-1 to 3. The adsorption isotherm models of LDHs-1 exhibit high R2 values which shown that their F− adsorption behaviour is the result of the combined effect of physical and chemical adsorption. The adsorption behaviour of LDHs-2 and 3 is more in line with the quasi second-order kinetic model and Langmuir model, indicating that the F− adsorption process is a single-layer adsorption based on chemical reaction-controlled kinetics. The lower temperatures are more conducive to the process for all three adsorbents adsorption process generates heat. FLDHs can still re-adsorb the dye CR after four cycles of adsorbing fluoride ions. HE-LDHs exhibit high adsorption, selectivity, and good reuse performance for fluoride ions, making them promising for wastewater treatment applications.

Creation of Composite Aerogels Consisting of Activated Carbon and Nanocellulose Blended with Cross-Linked Biopolymers: Application as Ethylene Scavengers

This study involved producing aerogels using activated carbon (AC) and nanocellulose (NC). Two distinct structured composites, AC composite aerogel (ACCA) and NC composite aerogel (NCCA), were developed by separately mixing AC and NC with identical proportions of cross-linked biopolymers: hydroxypropyl methylcellulose (HPMC), methylcellulose (MC), and chitosan (C). These aerogels were evaluated for their capability to adsorb ethylene gas through batch experiments, while the physical and chemical characteristics were thoroughly examined to determine their feasibility of removing ethylene. The resulting ACCA and NCCA aerogels exhibited low densities of 0.094 g cm−3 and 0.077 g cm−3, respectively, coupled with high porosity ranging between 95 and 96%. During the ethylene adsorption test, NCCA exhibited superior ethylene removal rates (~14.88–16.77 mL kg−1) compared to ACCA (~13.57–14.97 mL kg−1). Specifically, NCCA achieved a removal efficiency of 83.86% compared to 74.64% for ACCA. Kinetic model fitting yielded high R2 values ranging from 0.97 to 0.98 with the Lagergren kinetic model. These findings suggest the potential of composite aerogels to be incorporated into food packaging materials for dynamic ethylene capture, independent of environmental conditions, thereby providing promising routes for further development.

Determining a Meaningful R-squared Value in Clinical Medicine

Introduction In clinical medicine, determining a meaningful R2 value requires careful consideration of various factors. The R-squared (R2) value is a statistical measure used to assess the extent to which independent variables explain the dependent variable in regression models. Its interpretation in clinical medicine is very context-dependent and lacks a definitive threshold. While a higher R2 suggests a stronger relationship between variables, smaller R-squared values will still hold relevance, especially for multifactorial clinical outcomes. Due to clinical medicine’s genetic, environmental, and behavioral factors, high R2 values may not be realistic or expected in this context. Methods This study is a narrative review encompassing expert opinions, results from randomized controlled trials (RCTs), and observational studies relating to the use and interpretation of the coefficient of variance (R2) in clinical medicine. Results 30 studies spanning literature across various disciplines within clinical medicine were evaluated and synthesized to provide a contextualized, nuanced approach to interpreting the R2 in medical literature. Conclusions A nuanced understanding of the significance of R2 values in clinical medicine requires considering the research question, clinical context, complexity of the phenomenon, and contextual comparisons. Integrating statistical rigor and clinical judgment helps derive meaningful insights from clinical data. This paper highlights key considerations for evaluating a “good” R2 value in clinical medicine and concludes that an R2 of &gt;15% is a generally a meaningful value in clinical research.

A novel method to estimate the 3D chlorophyll a distribution in the South China Sea surface waters using hydrometeorological parameters

Chlorophyll a (Chl-a) is a key indicator of marine ecosystems, and certain hydro-meteorological parameters (HMPs) are highly correlated with its fluctuations. Here, relevant and accessible HMPs were used as inputs, combined with machine learning (ML) algorithms for estimating 3D Chl-a in the South China Sea (SCS). With the inputs of temperature, salinity, depth, wind speed, wind direction, sea surface pressure, and relative humidity, the LightGBM-based model performed well, achieving high R2 values of 0.985 and 0.789 in validation and testing sets, respectively. Based on a large number of in situ measurements, this model enables the estimation of the 3D distribution of summer Chl-a in the SCS over the past fifteen years using a 3D hydrographic dataset combined with surface meteorological parameters. The results show that the 3D distribution of the model estimated Chl-a is characterized similarly to the previous studies and can capture the effect of hydro-meteorological conditions on Chl-a distribution. The environmental variables affecting Chl-a were considered more comprehensively in this study, and the methodological framework has the potential to be applied to the low-cost monitoring of the remaining water quality parameters.

A smarter approach to liquefaction risk: harnessing dynamic cone penetration test data and machine learning for safer infrastructure

This paper presents a novel approach for assessing liquefaction potential by integrating Dynamic Cone Penetration Test (DCPT) data with advanced machine learning (ML) techniques. DCPT offers a cost-effective, rapid, and adaptable method for evaluating soil resistance, making it suitable for liquefaction assessment across diverse soil conditions. This study establishes a threshold criterion based on the ratio of the penetration rate to the dynamic resistance (e/qd), where values exceeding four indicate high liquefaction susceptibility. ML models, including Support Vector Machine (SVM) optimized with Particle Swarm Optimization (PSO), Grey Wolf Optimizer (GWO), Genetic Algorithm (GA), and Firefly Algorithm (FA), were employed to predict the e/qd ratio using key geotechnical parameters, such as fine content, peak ground acceleration, reduction factor, and penetration rate. The SVM-PSO model demonstrated superior performance, with high R2 values of 0.999 and 0.989 in the training and testing phases, respectively. The proposed methodology offers a sustainable and accurate approach for liquefaction assessment, reducing the environmental impact of geotechnical investigations, while ensuring reliable predictions. This study bridges the gap between field testing and advanced computational techniques, providing a powerful tool for geotechnical engineers to assess liquefaction risks and design resilient infrastructures.

Optimizing energy absorption and peak force in metal/glass fiber sandwich panels with trapezoidal cores

Sandwich panels with trapezoidal metal/glass fiber cores are increasingly popular due to their lightweight and energy-absorption properties. This study employs response surface methodology (RSM) and Box-Behnken design to investigate the effects of core angle, fiber orientation, and MCM-48 nanoparticles on the panels’ energy absorption and peak force, developing regression models with high R2 values of 0.9027 and 0.9228, respectively. Experimental tests were conducted to validate these models, showing minimal deviation from predicted values. Results indicate that increasing the fiber orientation angle from 30° to 90° enhances energy absorption and peak force by 72.18 and 46.9%, respectively, and adding MCM-48 nanoparticles up to 0.25% weight improves energy absorption by 60.8%. A core angle of 52° balances energy absorption and peak force, while integrating a metal wire mesh within the panels significantly enhances energy absorption and reduces core brittleness. The optimal parameters for maximum energy absorption and minimum peak force include a core angle of 58°, fiber orientation of 73.5°, and no nanoparticles. These findings provide valuable insights into the design and optimization of sandwich panels for various applications.

Integration of mathematical and experimental modeling for sustainable phycobiliprotein production via fed-batch cultures

The production of phycobiliproteins, such as cyanobacterial phycocyanins, is a growing interest due to their diverse industrial and biotechnological applications. This study focuses on optimizing phycocyanin production using the strain Potamosiphon sp. through experimental techniques and mathematical modeling in fed-batch cultures. The methodology applied includes determining the kinetic constants by linearizing the Monod equation evaluating the concentrations of biomass, C-phycocyanin (C-PC), nitrates (NO3), and phosphates (PO4). A mathematical model of periodic fed-batch feeding was subsequently established, applying mass conservation principles and evaluating the accuracy of the Monod, Contois, Moser, and Tessier models. The results indicate that phycocyanin production is highly dependent on phosphorus and nitrogen concentrations, with optimal conversion observed at specific levels of these elements (0.832 for phosphorus and 0.805 for nitrogen in terms of C-PC and biomass, respectively). The Tessier model demonstrated the highest accuracy in predicting production and optimizing operational conditions, with a Mean Squared Error (MSE) of 0.005000 for biomass production, 0.200000 for C-PC production, and 0.000010 for substrate consumption. It also achieved high R² values of 0.980 for biomass, 0.999 for C-PC production, and 0.997 for substrate consumption. It presented the lowest Akaike Information Criterion (AIC) scores, indicating its robustness and reliability in modeling these processes and manipulating cultivation conditions and providing adequate nutrition allowed for achieving growth rates of 1.23 g/L and a C-PC concentration of 37 mg/L, which are essential for industrial applications such as natural colorants and antioxidants, among others.

Ternary polymeric beads of chitosan-based polyaniline doped iron oxide used for removal of cadmium and lead ions from water: Kinetic, isotherm and mechanism studies

ABSTRACT In this study, the binary polymeric magnetic beads were synthesized successfully by one-pot polymerization of polyaniline (PA) doped magnetic iron oxide nanoparticles (MNPs) followed by chitosan (CTS) cross-linkage. The newly fabricated ternary magnetic beads (CTS@MPA) were applied to enhance the removal of Pb(II) and Cd(II) ions from the aqueous media. The prepared nanocomposite was characterized using Fourier-transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDX), and vibrating sample magnetometry (VSM). Adsorption experiments were carried out under optimum conditions. To evaluate the experimental data, the isotherm and kinetic models were conducted. The behavior of adsorption was assessed, and the coefficient of determination values demonstrated that the adsorption process best fitted the Freundlich model due to high values of R2 (0.957 & 0.981). The maximum adsorption capacity was obtained at 150.82 mg.g−1 and 169.21 mg.g−1 for Pb(II) and Cd(II) ions, respectively. Kinetic studies showed that adsorption follows a pseudo-second-order model (R2 = 0.952 & 0.985). Weber-Morris claims that intraparticle diffusion plays a significant role in the adsorption kinetics. Finally, the thermodynamic study is demonstrated, and the adsorption is endothermic, and the mechanism follows a physisorption pattern.

Repeatability and Reproducibility of pseudo-Continuous Arterial Spin Labeling Measured Brain Perfusion in Healthy Volunteers and Glioblastoma Patients.

Arterial spin labeled (ASL) MRI has gained recognition as a quantitative perfusion imaging method for managing patients with brain tumors. Limited studies have so far investigated the reproducibility of ASL-derived perfusion in patients with brain tumors. This study aims to evaluate intrasession repeatability and intersession reproducibility of perfusion measurements using 3D pseudo-continuous ASL (pCASL) with Cartesian TSE (TSE-CASPR) in healthy volunteers (HV) and glioblastoma (GBM) patients at 3 Tesla and compare against 3D pCASL with GRASE. This prospective study (NCT03922984) was approved by the institutional review board and written informed consent was obtained from all subjects. HV underwent repeated pCASL evaluations 2-4 weeks apart between November 2021 and October 2022. GBM patients were recruited for longitudinal MRI from September 2019 to February 2023. Intrasession repeatability (HV and GBM) and intersession reproducibility (HV only) of pCASL were assessed using linear regression, Bland-Altman analyses, intraclass correlation coefficient (ICC) with 95% confidence interval (CI), and within-subject coefficients of variation (wsCV). Twenty HV (9 men, age: 25.1±1.7 years, range 23-30 years) and 21 GBM patients (15 men; age: 59.8±14.3 years, range 28-81 years) were enrolled. Within imaging session, 3D pCASL measured perfusion with TSE-CASPR and GRASE respectively achieved high R2 values (0.88-0.95; 0.93-0.96), minimal biases (-0.46 to 0.81; -0.08 to 0.35 mL/100g/min), high ICC [95% CI] (0.96-0.98 [0.94-0.98]; 0.96-0.98 [0.92-0.99]), and low wsCV (6.64%-9.07%; 5.20%-8.16%) in HV (N=20) and GBM patients (N=21). Across imaging session, 3D pCASL in HV (N=20) achieved high R2 values (0.71; 0.82), minimal biases (-1.2; -0.90 mL/100g/min), high ICC [95% CI] values (0.85 [0.81-0.89]; 0.90 [0.87-0.93]), and low wsCV values (13.82%; 9.98%). Our study demonstrated excellent intrasession repeatability of 3D pCASL measured cerebral perfusion in HV and GBM patients and good to excellent intersession reproducibility in HV. 3D pCASL with GRASE performed slightly better than 3D pCASL with TSE-CASPR in HV; however, in GBM patients, 3D pCASL with TSE-CASPR showed better performance in tumor regions with nearly twofold higher SNR. ASL measured perfusion could serve as a non-contrast quantitative imaging biomarker to facilitate the management of GBM patients. ASL = arterial spin labeling; pCASL = pseudo-continuous arterial spin labeling; GBM = glioblastoma; CBF = cerebral blood flow; CASPR = Cartesian acquisition with spiral profile reordering; GRASE = gradient and spin echo; NAGM = normal-appearing gray matter.

Experimental study and mathematical modeling of drying kinetics of mint leaves under forced convection

In this study, we investigated the influence of temperature and air velocity on the drying behavior of Mentha spicata L. leaves using a forced convection dryer. This drying method was chosen for its cost-effectiveness and its ability to facilitate fast and efficient drying while preserving the quality of aromatic herbs. Our aim was to identify the most suitable mathematical model for accurately representing the drying process. We analyzed experimental data on moisture content obtained during thin-layer drying of Mentha spicata L. using various established empirical models from existing literature. Thirteen empirical models were applied to fit the drying curves obtained. Statistical analysis, employing key parameters such as the coefficient of determination (R2), reduced chi-square (χ2), and root mean square error (RMSE), indicated that the Midilli and Kucuk model provided the most accurate fit for the observed drying curves. This was evidenced by the highest R2 values (0.9994, 1.0000), lowest χ2 values (6.07 × 10−6, 6.28 × 10−5), and lowest RMSE values (2.17 × 10−3, 6.86 × 10−3). We observed an increasing trend in moisture diffusivity with temperature variations between 32 °C and 60 °C, while maintaining a constant airflow velocity of 1.2 m/s. Additionally, an increase in mean air velocity from 0.7 to 1.7 m/s correlated with higher moisture diffusivity. The effectiveness of the Arrhenius equation in representing the temperature dependence of diffusivity was confirmed, with an activation energy of 72 kJ/mol, consistent with previous literature findings.

Investigation of the relationship between socioeconomic status and literacy in PISA Türkiye data

Previous researchers have identified socioeconomic status as a significant predictor of achievement/literacy. However, it is important to recognize that the influence of socioeconomic status on literacy may vary at different levels of socioeconomic status. Thus, this study analyzes the relationship between socioeconomic status and literacy scores for all domains in PISA Türkiye data from 2003 to 2022 through the Classification and Regression Trees and linear regression methods. Upon examining the results, separate investigations carried out for the lower and upper socioeconomic status groups indicate that R2 values were found to be equal to or greater than .80 in 37 out of the 42 analyses. From 2003 to 2009, the R2 values in both groups were considerably high; however, there has been a notable decline in subsequent periods. The year 2009 demonstrated particularly high R2 values by ESCS in all domains for both upper and lower groups. Consequently, socioeconomic status exhibited a greater predictive power on literacy scores across all domains in the lower socioeconomic group than upper socioeconomic group.

Sensitivity and Uncertainty Analysis of the GeeSEBAL Model Using High-Resolution Remote-Sensing Data and Global Flux Site Data

Actual evapotranspiration (ETa) is an important component of the surface water cycle. The geeSEBAL model is increasingly being used to estimate ETa using high-resolution remote-sensing data (Landsat 4/5/7/8). However, due to surface heterogeneity, there is significant uncertainty. By optimizing the quantile values of the reverse-modelling automatic calibration algorithm (CIMEC) endpoint-component selection algorithm under extreme conditions through 212 global flux sites, we obtained the optimized quantile values of 11 vegetation types of cold- and hot-pixel endpoint components (Ts and NDVI). Based on the observation data of the global FLUXNET tower, the sensitivity of 20 parameters in the improved geeSEBAL model was determined through Sobol’s sensitivity analysis. Among them, the parameters dT and SAVI,hot were confirmed as the most sensitive parameters of the algorithm. Subsequently, we used the differential evolution Markov chain (DE-MC) method to analyse the uncertainty of the parameters in the geeSEBAL model used the posterior distribution of the parameters to modify the sensitive parameter values or ranges in the improved geeSEBAL model and to simulate the daily ETa. The results indicate that by analysing the end element components of the geeSEBAL model (Ts and NDVI), quantile numerical optimization and parameter optimization can be performed. Compared with the original algorithm, the improved geeSEBAL model has significantly improved simulation performance, as shown by higher R2 values, higher NSE values, smaller bias values, and lower RMSE values. The most suitable values of the predefined parameter Zoh were determined, and the reanalysis of meteorological data inputs (relative humidity (RH), temperature (T), wind speed (WS), and net radiation (Rn)) was also found to be an important source of uncertainty for the accurate estimation of ETa. This study indicates that optimizing the quantiles and key parameters of the model end component has certain potential for further improving the accuracy of the geeSEBAL model based on high-resolution remote-sensing data in estimating the ETa for various vegetation types.

Estimation on compressive strength of recycled aggregate self-compacting concrete using interpretable machine learning-based models

PurposeRecycled aggregate self-compacting concrete (RASCC) has the potential for sustainable resource utilization and has been widely applied. Predicting the compressive strength (CS) of RASCC is challenging due to its complex composite nature and nonlinear behavior.Design/methodology/approachThis study comprehensively evaluated commonly used machine learning (ML) techniques, including artificial neural networks (ANN), random trees (RT), bagging and random forests (RF) for predicting the CS of RASCC. The results indicate that RF and ANN models typically have advantages with higher R2 values, lower root mean square error (RMSE), mean square error (MSE) and mean absolute error (MAE) values.FindingsThe combination of ML and Shapley additive explanation (SHAP) interpretable algorithms provides physical rationality, allowing engineers to adjust the proportion based on parameter analysis to predict and design RASCC. The sensitivity analysis of the ML model indicates that ANN’s interpretation ability is weaker than tree-based algorithms (RT, BG and RF). ML regression technology has high accuracy, good interpretability and great potential for predicting the CS of RASCC.Originality/valueML regression technology has high accuracy, good interpretability and great potential for predicting the CS of RASCC.

International Air Cargo Volume Forecasting based on the Stacking Fusion Model

Accurate forecasting of international air cargo volumes is crucial for air cargo carriers, which optimizes operational efficiency and profitability and strengthens market competitiveness and risk response capabilities. This study focuses on the open-source cargo data of over 100 international airlines in India from January 2015 to March 2017. Firstly, the missing data must be dealt with. Then, histograms and probability density plots are used to depict the distribution of the cargo volume and analyze the dynamics of the cargo volume of each company at annual and quarterly levels. Subsequently, key factors affecting cargo volumes are analyzed, with a special focus on the cargo performance of the Top 5 airlines, which is visualized in box plots. Given the time-series nature of cargo data, this study adopts the data sequential partitioning method. It innovatively utilizes the Stacking ensemble learning framework, integrating regression models such as Random Forest, Decision Tree, Extreme Gradient Boosting, and K-Nearest Neighbor as the base learners. The meta-learner is constructed to integrate the outputs of each base model, and the hyperparameters of the model are optimized by grid search and cross-validation. Comparison results show that the Stacking model has a higher R² value and lower MSE than any single base learners, indicating that it is more effective in international air cargo volume prediction. The prediction model based on stacking ensemble learning provides an effective strategy for improving air cargo prediction accuracy.