High R2 Values Research Articles

Effects of CO2 concentration and time on algal biomass film, NO3–N concentration, and pH in the membrane bioreactor: Simulation-based ANN, RSM and NSGA-II

The practice of aquaculture is associated with the generation of a substantial quantity of effluent. Microalgae must effectively assimilate nitrogen and phosphorus from their surrounding environment for growth. This study modeled the algal biomass film, NO3–N concentration, and pH in the membrane bioreactor using the response surface methodology (RSM) and an artificial neural network (ANN). Furthermore, it was suggested that the optimal condition for each parameter be determined. The results of ANN modeling showed that ANN with a structure of 5-3 and employing the transfer functions tansig-logsig demonstrated the highest level of accuracy. This was evidenced by the obtained values of coefficient (R2) = 0.998, R = 0.999, mean squared error (MAE) = 0.0856, and mean square error (MSE) = 0.143. The ANN model, characterized by a 5-5 structure and employing the tansig-logsig transfer function, demonstrates superior accuracy when predicting the concentration of NO3–N and pH. This is evidenced by the high values of R2 (0.996), R (0.998), MAE (0.00162), and MSE (0.0262). The RSM was afterward employed to maximize the performance of algal film biomass, pH levels, and NO3–N concentrations. The optimal conditions for the algal biomass film were a concentration of 2.884 mg/L and a duration of 6.589 days. Similarly, the most favorable conditions for the NO3–N concentration and pH were 2.984 mg/L and 6.787 days, respectively. Therefore, this research uses non-dominated sorting genetic algorithm II (NSGA II) to find the optimal NO3–N concentration, algal biomass film, and pH for product or process quality. The region has the greatest alkaline pH and lowest NO3–N content.

Optimization of Methylene Blue Adsorption onto Activated Carbon derived from Pineapple Peel Waste using Response Surface Methodology

This study focuses on the optimization of methylene blue (MB) adsorption onto activated carbon derived from pineapple peel waste (ACPPW) using response surface methodology (RSM). The activated carbon was synthesized via a chemical activation method and characterized using FTIR spectroscopy. The adsorption process was optimized by investigating the effects of three key parameters: adsorbent dosage, initial dye concentration, and contact time. A Central Composite Design (CCD) with 20 runs was employed, and ANOVA was performed to determine the significance of each parameter. The central composite design (CCD) was employed to determine the effect of three independent variables, namely adsorbent dosage (X1), initial MB concentration (X2), and contact time (X3), on the response variable, MB removal efficiency (Y). The optimal conditions for achieving a high adsorption capacity of 98.19 % for methylene blue dye were determined to be an adsorbent dosage of 400 mg/L, an initial dye concentration of 15 mg/L, and a contact time of 12 minutes. The regression equation achieved a high R2 value of 0.9883, indicating a good fit for the experimental data and the model's reliability. These findings provide valuable insights into efficiently utilizing pineapple peel waste as a sustainable source for producing activated carbon with excellent adsorption capabilities for dye removal, thus contributing to environmental sustainability and waste management efforts.

Chemical and Thermal Treatment for Drying Cassava Tubers: Optimization, Microstructure, and Dehydration Kinetics.

Perishable commodities like cassava necessitate effective postharvest preservation for various industrial applications. Hence, optimizing pretreatment processes and modeling drying kinetics hold paramount importance. This study aimed to optimize cassava pretreatment using the central composite design of a response surface methodology while also assessing microstructure and dehydration kinetics. Diverse chemical and thermal pretreatments were explored, encompassing sodium metabisulfite concentrations (0-4% w/w), citric acid concentrations (0-4% w/w), and blanching time (0-4 min). The four investigated responses were moisture content, whiteness index, activation energy (Ea), and effective moisture diffusivity (Deff). Employing five established drying models, suitability was appraised after optimal pretreatment conditions were determined. The findings revealed that moisture content ranged from 5.82 to 9.42% db, whereas the whiteness index ranged from 87.16 to 94.23. Deff and Ea ranged from 5.06 × 10-9 to 6.71 × 10-9 m2/s and 29.65-33.28 kJ/mol, respectively. The optimal pretreatment conditions for dried cassava were identified by optimizing the use of 1.31% citric acid, 1.03% sodium metabisulfite, and blanching time for 1.01 min. The microstructure indicated that particular chemical and thermal pretreatment configurations yielded particles in the shape of circular and elliptical granules. The logarithmic model provided the most accurate description of the dehydration kinetics, with the highest R2 value (0.9859) and the lowest χ2, RSME, and SSE values of 0.0351, 0.0015, and 0.0123, respectively.

Accurate prediction of pressure losses using machine learning for the pipeline transportation of emulsions

One of the significant challenges to designing an emulsion transportation system is predicting frictional pressure losses with confidence. The state-of-the-art method for enhancing reliability in prediction is to employ artificial intelligence (AI) based on various machine learning (ML) tools. Six traditional and tree-based ML algorithms were analyzed for the prediction in the current study. A rigorous feature importance study using RFECV method and relevant statistical analysis was conducted to identify the parameters that significantly contributed to the prediction. Among 16 input variables, the fluid velocity, mass flow rate, and pipe diameter were evaluated as the top predictors to estimate the frictional pressure losses. The significance of the contributing parameters was further validated by estimation error trend analyses. A comprehensive assessment of the regression models demonstrated an ensemble of the top three regressors to excel over all other ML and theoretical models. The ensemble regressor showcased exceptional performance, as evidenced by its high R2 value of 99.7 % and an AUC-ROC score of 98 %. These results were statistically significant, as there was a noticeable difference (within a 95 % confidence interval) compared to the estimations of the three base models. In terms of estimation error, the ensemble model outperformed the top base regressor by demonstrating improvements of 6.6 %, 11.1 %, and 12.75 % for the RMSE, MAE, and CV_MSE evaluation metrics, respectively. The precise and robust estimations achieved by the best regression model in this study further highlight the effectiveness of AI in the field of pipeline engineering.

Impact of extrusion processing on bioactive compound enriched plant-based extrudates: A comprehensive study and optimization using RSM and ANN-GA

The present study aimed to investigate the influence of extrusion process parameters, specifically screw speed (300 to 500 rpm) and temperature (130 to 150 °C) on the stability of phytosterols and the physical, functional, and structural characteristics of extrudates made from composite blends containing pea protein isolate (PPI). Response Surface Methodology (RSM) and Artificial Neural Network-Genetic Algorithm (ANN-GA) were employed to optimize the process parameters.Extrudates produced at lower temperatures exhibited higher phytosterol retention compared to those produced at higher temperatures. X-ray diffraction analysis showed that incorporating PPI significantly increased crystallinity from 22% to 33%. Fourier Transform Infrared analysis revealed a reduction in the Amide II stretching of the secondary structure in pea protein isolate following extrusion. Scanning electron microscopy images demonstrated that the addition of PPI and functional oil altered the structure of the extrudates.Statistical comparison of the process parameters using higher R2 values, lower root mean square as well as mean absolute percentage error indicated that the ANN model outperformed the RSM model in predicting all responses. Overall, incorporating PPI and phytosterols into corn-based extrudates improved their nutritional quality. These findings contribute to meeting the demand for fortified foods.

Development of a Heat Index Related to Air Quality and Meteorology for an Assessment of Work Performance in Thailand’s Urban Areas

A heat index is a key indicator directly related to meteorological factors influencing human health, particularly work performance. However, the interaction between air quality, meteorology, heat, and associated work performance is loosely defined, especially in urban areas. In this study, we develop a heat index (HI) related to air quality terms, including PM2.5, NOx, and CO, and meteorology terms, including temperature and relative humidity, to assess work performance in Thailand’s urban areas, including Chiang Mai, Bangkok, Nakhon Ratchasima, and Ubon Ratchathani, using a multivariate regression model. The regression models’ performance shows high R2 values ranging from 0.82 to 0.97, indicating a good level of performance. A recurring trend across all locations is elevated HI values during April and May, signifying typical pre-monsoon conditions in tropical regions. Following this peak, the values of the heat index (HI) begin to fall, possibly due to the start of the wet season. As shown by the decrease in productivity during periods of elevated heat index values, the observed increase in temperatures has noticeable effects on work performance.

INDIVIDUAL TREE-BASED FOREST SPECIES DIVERSITY ESTIMATION USING UAV-BORNE HYPERSPECTRAL AND LIDAR DATA

Abstract. Forest biodiversity is essential in maintaining ecosystem functions and services. Recently, unmanned aerial vehicle (UAV) remote sensing technology has emerged as a cost-effective and flexible tool for biodiversity monitoring. In this study, we compared the optimal clustering algorithm, classification method (spectral angle mapper, SAM), spectral diversity metric and structural heterogeneity index for forest species diversity estimation in two complex subtropical forests, Mazongling (MZL) and Gonggashan (GGS) National Nature Forest Reserves in China, using UAV-borne hyperspectral and LiDAR data. The results showed that the SAM classification method performed better with higher values of R2 than the clustering algorithm for predicting both species richness (MZL: 0.62 > 0.46 and GGS: 0.55 > 0.46) and Shannon-Wiener index (MZL: 0.64 > 0.58 and GGS: 0.52 > 0.47), while the optimal clustering algorithm had the highest prediction accuracy for the Simpson index, followed by the SAM classification method, spectral diversity metric and structural heterogeneity index (MZL: 0.83>0.44>0.31>0.12, GGS: 0.62>0.44>0.38>0.00). Our study indicated that the SAM classification method had the advantage of identifying rare species and estimating species richness, while the clustering method could capture forest diversity patterns rapidly without distinguishing the specific tree species and predict the Simpson index more accurately. Overall, both clustering and classification methods exhibited superior performance compared to spectral or structural diversity indices. Our findings highlight the applicability of UAV remote sensing in monitoring forest species diversity in complex subtropical forests.

Synthesis, Characterization, and Performance of Nano-Metal-Oxide (Al2O3) Blended Biochar for the Removal of Iron from Contaminated Water for Enhanced Kinetic and Adsorption Studies

This paper explored synthesis, characterization, and adsorption modeling for the application of nano-metal-oxide (Al2O3) blended biochar (NMOBC) derived from date palm waste in removing iron (Fe3+) from contaminated water. The pseudo-second-order model provided a goodness-of-fit that was superior to the pseudo-first-order kinetic model based on the value of R2 with all of the initial concentrations. The Elovich kinetic model also presented a good fit, indicating that chemisorption is a predominant mechanism in the adsorption process. The Langmuir, Freundlich, Redlich-Petersons, Temkin, and Sips models provided an exclusive perspective of the adsorption dynamics based on the high value of R2. However, the Sips model suggested the best fit of all of the employed models, with the lowest RMSE value of 0.0239 mg/g and the maximum adsorption capacity of 22.680 mg/g for NMOBC. Both adsorbents were effectively regenerated and reused in multiple cycles, thus leading to sustainable practices. Numerous analytical techniques, including SEM/EDX, FTIR, and BET, were employed in characterizing the structural, morphological, and functional properties of the synthesized NMOBC and BC. Subsequently, it revealed that the adsorption process and the role of various interactions are attributed to surface area, porosity, and ion exchange.

Industrial-scale anaerobic Co-digestion (ACoD) of palm oil mill effluent (POME) and decanter cake (DC) for maximizing methane yield: An integrated machine learning and simulation-based economic analysis approach

This study employs machine learning (ML) algorithms, including multiple linear regression (MLR), decision tree (DT) and support vector regression (SVR) to predict the methane yield from the anaerobic co-digestion (ACoD) of palm oil mill effluent (POME) and decanter cake (DC) in an industry-scale anaerobic covered lagoon. Results showed that the DT model outperformed the other models with a high R2 value of 0.9763 and the lowest MAE, MSE, and RMSE values. It exhibited over 95 % similarity to the actual results, validating its effectiveness in capturing real-world scenarios. Optimisation was conducted using response surface methodology (RSM) to achieve maximum biogas production (14,245 m3/day) and methane yield (0.285 Nm3 CH4/kg CODremoved), where the optimal range of pH, organic loading rate (OLR) and dilution ratio of DC were found to be 6.83–6.94, 0.82–0.83 kg COD/m3. day and 0.09–0.11 respectively. The ACoD process was simulated, and an economic analysis was performed using SuperPro Designer v10.6. ACoD of POME and DC was more economically viable than mono-digestion of POME, with a 43.16 % improvement in return on investment (ROI), considering the trade-off between the additional cost of pre-treating DC and the additional revenue from improved biogas production using ACoD.

Integrated computational approaches for designing potent pyrimidine-based CDK9 inhibitors: 3D-QSAR, docking, and molecular dynamics simulations

CDK9 is an emerging target for the development of anticancer drugs. The development of CDK9 inhibitors with significant potency had consistently posed a formidable challenge. In the current research, a number of computational methodologies, such as, 3D-QSAR, molecular docking, fingerprint analysis, molecular dynamic (MD) simulations followed by MMGB/PBSA and ADMET studies were used systemically to uncover the binding mechanism of pyrimidine derivatives against CDK9. The CoMFA and CoMSIA models having high q2 (0.53, 0.54) and r2 values (0.96, 0.93) respectively indicating that model could accurately predict the bioactivities of CDK9 inhibitors. Using the R-group exploration technique implemented by the Spark™ by Cresset group, the structural requirements revealed by the contour maps of model were utilized strategically to create an in-house library of 100 new CDK9 inhibitors. Additionally, the compounds from the in-house library were mapped into 3D-QSAR model which predicted pIC50 values comparable to the experimental values. A comparison between 3D-QSAR generated contours and molecular docking conformation of ligands was performed to elucidate the essentials of CDK9 inhibitor design. MD simulations (100 ns) were performed on the selected docked complexes A21, A14 and D98 which contributed in validating the binding interactions. According to the findings of binding free energy analysis (MMGB/PBSA), It was observed that residues CYS106 and GLU107 had a considerable tendency to facilitate ligand-protein interactions via H-bond interactions. The aforementioned findings have the potential to enhance researchers comprehension of the mechanism underlying CDK9 inhibition and may be utilized in the development of innovative and efficacious CDK9 inhibitors.

Plant-based monitoring techniques to detect yield and physiological responses in water-stressed pepper

Today, the use of sensors and imaging techniques, which are used to obtain information about plants and soil in smart irrigation systems, is rapidly becoming widespread. This study aimed to investigate the usability of leaf turgor pressure and thermal images from plant-based monitoring techniques to detect water stress and the irrigation time of pepper (Capsicum annuum L. cv. "California Wonder") and to determine their relationship with physiological traits in Canakkale/Türkiye in 2017 and 2018. The four irrigation treatments (100%, 75%, 50%, and 25%) were applied in the experiment. Leaf turgor pressure (Pp), thermal images and physiological measurements were carried out during the growing season. Soil moisture and Pp were monitored in real time by remote. Thermal and physiological measurements were made before each irrigation. As a result of the study, the average evapotranspiration (ETc) was 697 mm, and the yield value was 83.7 t ha−1 under non-stress conditions. Depending on the decrease in ETc, yield values also decreased significantly. Leaf water potential and stomatal conductivity values were statistically different in all irrigation treatments. The change in the activity of catalase (CAT) due to water stress was greater than that of superoxide dismutase (SOD). In this case, it can be said that other physiological traits are more successful than SOD in distinguishing water stress. According to the regression models, significant relationships were determined between both the indices calculated from the thermal images and Pp, yield, and physiological traits. The predictive ability of Pp values has been strengthened with the addition of meteorological properties to the model in general. The highest correlation (R2 =0.63) was between Pp + meteorological properties and CAT. All the regression models between physiological traits and indices calculated from thermal images were statistically significant. The highest R2 values were obtained in August. In this month, the highest correlations were between Crop Water Stress Index (CWSIp) and leaf water potential / stomatal conductivity (R2 =0.91), IGp and stomatal conductivity (R2 =0.80). The predictive power of CWSIp was higher than Stomatal Conductivity Index (IGp). The experiment illustrated that Pp and temperature data, which are plant-based monitoring methods, have the potential to detect water stress in peppers.

Optimization of operational parameters using RSM, ANN, and SVM in membrane integrated with rotating biological contactor

Membrane fouling is a critical bottleneck to the widespread adoption of membrane separation processes. It diminishes the membrane permeability and results in high operational energy costs. The current study presents optimizing the operating parameters of a novel rotating biological contactor (RBC) integrated with an external membrane (RBC + ME) that combines membrane technology with an RBC. In the RBC + ME, the membrane panel is placed external to the bioreactor. Response surface methodology (RSM) is applied to optimize the membrane permeability through three operating parameters (hydraulic retention time (HRT), rotational disk speed, and sludge retention time (SRT)). The artificial neural networks (ANN) and support vector machine (SVM) are implemented to depict the statistical modelling approach using experimental data sets. The results showed that all three operating parameters contribute significantly to the performance of the bioreactor. RSM revealed an optimum value of 40.7 rpm disk rotational speed, 18 h HRT and 12.4 d SRT, respectively. An ANN model with ten hidden layers provides the highest R2 value, while the SVM model with the Bayesian optimizer provides the highest R2. RSM, ANN, and SVM models reveal the highest R-square values of 0.97, 0.99, and 0.99, respectively. Machine learning techniques help predict the model based on the experimental results and training data sets.

Mathematical Modeling and Validation of Solar-Induced Ventilation System for Vehicle Cabin Cooling in Hot Parking Conditions

Exposure to direct sunlight raises interior temperatures in vehicle cabins, risking heat-related illnesses. Solar passive techniques mitigate this issue, especially during hot parking conditions. However, a comprehensive mathematical model encompassing both solar chimney and vehicle cabin is lacking. This study develops a novel mathematical model to predict temperature distribution and airflow rate, enhancing system performance evaluation. The system comprises a solar air collector with an adjustable arm mounted on the vehicle's roof. The model was validated theoretically and experimentally. The experimental work is conducted with the physical model with an air gap of 0.1 m, 0.77 m width, and a 1.12 m collector length under outdoor conditions. Results indicated a gradual increase in temperatures of the glass cover, air in the collector channel, absorber, and mass airflow rate with solar radiation intensity, significantly influencing system performance. The high value of R2 and the consistency of the model's results with theoretical and experimental outcomes justified the validity and accuracy of the proposed model, exhibiting a deviation percentage of less than 10%. The developed model can be utilized to study influential parameters for optimizing the proposed strategy's performance and components, yielding comparable results to experimental data. Additionally, it provides researchers and car makers with a broader perspective and a range of options for further improvement in weight, size, cost, and aerodynamic of the vehicle.

Thermal optimization of MHD nanofluid over a wedge by using response surface methodology: Sensitivity analysis

It is well documented that heat transfer is enhanced with addition of nanosized particles in fluid. But, in a mechanical system there are variety of factors influences the heat transfer. Some factors are significant while others are not. In this paper, authors will discuss sensitivity of different input parameters such as Le, Nt and Nb on output responses Nux and Shx. To achieve this goal, the problem is modeled using basic conservation laws. The formulated model is a set of PDEs, which are converted to set of non-linear ODEs by using similarity transformation. Then these ODEs are solved numerically by using MATLAB built in package bvp4c and compared the numerical results with existing work and found good results. Sensitivity analysis is performed by employing RSM to determine the relationship between the input parameters such that 0.1≤Le≤1, 0.1≤Nt≤1 and 0.1≤Nb≤1 and the output responses (Nux and Shx). ANOVA tables are generated by using RSM. By using the ANOVA tables the correlations between input parameters and output response are developed. To check the validity of correlated equations, the residuals are plotted graphically and show best correlations between input parameters and output responses. The high values of R2=98.65 and AdjR2=97.43 for Nux and R2=97.83 and AdjR2=95.88 for Shx demonstrates the high validity of ANOVA results to perform sensitivity analysis. Finally, we have conducted a sensitivity analysis of the responses and came to the important results that Nt and Nb is most sensitive to Nusselt number and Sherwood number respectively.

Optimizing biochar-based geopolymer composites for enhanced water resistance in asphalt mixes: an experimental, microstructural, and multi-objective analysis

Due to increased traffic and environmental concerns, this study addresses challenges in conventional asphalt concrete. Our focus is on enhancing the water resistance of asphalt mixes through the optimization of both the asphalt binder and the biochar-based geopolymer composite. We employ experiments and response surface methodology to assess their impact on volume, Marshall parameters, and water resistance. Asphalt binders were evaluated within the range of 4–6%, while biochar-based geopolymer composite levels varied from 0 to 4%. According to the findings, the incorporation of the biochar-based geopolymer composite improves asphalt properties, stiffness, and temperature sensitivity. Response surface methodology (RSM) was utilized to construct robust mathematical models with high R2 values (90%) and low p-values. Multi-objective optimization indicated that optimal content levels were 4.56% for the binder and 2.71% for the biochar-based geopolymer composite. Model accuracy was confirmed with less than a 5% error in validation tests. The research also identified structural changes in the asphalt binder caused by the BGC Si–O phase. Additionally, the leaching value for both BGC and BGC-MAB asphalt concrete was found to be substantially below the regulatory limit, demonstrating the environmental safety of incorporating BGC into the asphalt sector.

Surface Soil Moisture Determination of Irrigated and Drained Agricultural Lands with the OPTRAM Method and Sentinel-2 Observations

Surface soil moisture (SSM) is one of the factors affecting plant growth. Methods involving direct soil moisture measurement in the field or requiring laboratory tests are commonly used. These methods, however, are laborious and time-consuming and often give only point-by-point results. In contrast, SSM can vary across a field due to uneven precipitation, soil variability, etc. An alternative is using satellite data, for example, optical data from Sentinel-2 (S2). The main objective of this study was to assess the accuracy of SSM determination based on S2 data versus standard measurement techniques in three different agricultural areas (with irrigation and drainage systems). In the field, we measured SSM manually using non-destructive techniques. Based on S2 data, we estimated SSM using the optical trapezoid model (OPTRAM) and calculated eighteen vegetation indices. Using the OPTRAM model gave a high SSM estimating accuracy (R2 = 0.67, RMSE = 0.06). The use of soil porosity in the OPTRAM model significantly improved the results. Among the vegetation indices, at the NDVI ≤ 0.2, the highest value of R2 was obtained for the STR to OPTRAM index, while at the NDVI > 0.2, the shadow index had the highest R2 comparable with OPTRAM.

Evaluation of Infiltration Modeling in the Cisadane Watershed in Indonesia: Existing and New Approach Equation

This study aims to evaluate the accuracy of semi-empirical models for determining infiltration based on the existing equations (Horton, Kostiakov and Kostiakov–Lewis) and a new form equation with optimized parameters which has good accuracy (lowest RMSE and MBE), sufficient linearity (highest R2), and a Willmott index close to 1. The location of this study was the Cisadane watershed, which is a priority watershed in Indonesia. The lowest values of average RMSE and MBE and highest values of R2 and Willmott index are given by Equation 1, which indicated that the model was closest to the field infiltration conditions. The distribution of the RMSE values for all survey points using Equation 1 resulted in small RMSE values ranging from 0.09 to 1.83. The distribution of R2 values for all survey points using Equation 1 yielded R2 values close to 1 (high linearity) with values ranging from 0.96 to 1. Based on these results, it can be concluded that Equation 1 is the most accurate infiltration model for the Cisadane watershed compared to the existing infiltration rate equation. This research can be used as an initial idea to create the infiltration rate model using other forms of equations with optimization parameters to produce a more representative model with field data.

Optimization of Heat Pump Dryer Conditions on Bioaccessibility of Some Secondary Metabolites of Cornelian Cherry-Capia Pepper Pestil

Pestil is a traditional and functional snack produced by removing moisture from fruit pulp using different methods. Although solar drying is common in traditional pestil drying, modern drying methods are preferred in industrial production. As one of these methods, heat pump drying (HPD) is considered to be the most favorable technique owing to its notable efficiency in the drying process and low energy consumption. Additionally, it has the capability to accurately regulate drying conditions, thereby enhancing the overall quality of the dried foods. The aim of this study was to investigate the effects of HPD conditions on drying kinetics and quality profile (total monomeric anthocyanin (TMA) content, total phenolic content (TPC), antioxidant capacity (AC), and carotenoid content) and their in vitro bioaccessibility of the cornelian cherry-capia pepper pestil. Also, it was aimed to produce high-quality pestil having bioactive properties by optimizing HPD conditions. Drying temperature (40–50°C) and cornelian cherry concentration (30–40%) were selected as independent variables for the response surface methodology (RSM). After in vitro digestion, pestil samples have higher bioaccessible β-carotene, α-carotene, lutein, and AC (especially for the DPPH method), whereas the drying process reduced the bioaccessibility of TPC in pestil samples. Using a RSM, we found that the way the responses (TMA, TPC, AC, and carotenoids) were related to the independent variables can be best explained by quadratic (Qc), reduced quadratic (RQc), and reduced cubic (RCc) models. These models had high R2 values, which mean they can accurately predict the outcomes. The optimal condition of responses with composite desirability of 0.852 was drying temperature at 46.68°C and 44.94% cornelian cherry concentration.

Evaluating Familiarity Ratings of Domain Concepts with Interpretable Machine Learning: A Comparative Study

Psycholinguistic properties such as concept familiarity and concreteness have been investigated in relation to technological innovations in teaching and learning. Due to ongoing advances in semantic representation and machine learning technologies, the automatic extrapolation of lexical psycholinguistic properties has received increased attention across a number of disciplines in recent years. However, little attention has been paid to the reliable and interpretable assessment of familiarity ratings for domain concepts. To address this gap, we present a regression model grounded in advanced natural language processing and interpretable machine learning techniques that can predict domain concepts’ familiarity ratings based on their lexical features. Each domain concept is represented at both the orthographic–phonological level and semantic level by means of pretrained word embedding models. Then, we compare the performance of six tree-based regression models (adaptive boosting, gradient boosting, extreme gradient boosting, a light gradient boosting machine, categorical boosting, and a random forest) on domain concepts’ familiarity rating prediction. Experimental results show that categorical boosting with the lowest MAPE (0.09) and the highest R2 value (0.02) is best suited to predicting domain concepts’ familiarity. Experimental results also revealed the prospect of integrating tree-based regression models and interpretable machine learning techniques to expand psycholinguistic resources. Specifically, findings showed that the semantic information of raw words and parts of speech in domain concepts are reliable indicators when predicting familiarity ratings. Our study underlines the importance of leveraging domain concepts’ familiarity ratings; future research should aim to improve familiarity extrapolation methods. Scholars should also investigate the correlation between students’ engagement in online discussions and their familiarity with domain concepts.

Prediction of water distribution uniformity of sprinkler irrigation system based on machine learning algorithms

The coefficients of uniformity Christiansen's uniformity coefficient (CU) and distribution uniformity (DU) are an important parameter for designing irrigation systems, and are an accurate measure for water lose. In this study, three machine learning algorithms Random forest (RF), extreme gradient boosting (XGB) and random forest-extreme gradient boosting (XGB-RF) were developed to predict the water distribution uniformity based on operating pressure, heights of sprinkler, discharge, nozzle diameter, wind speed, humidity, highest and lowest temperature for three different impact sprinklers (KA-4, FOX and 2520) for square and triangular system layout based on four scenarios (input combinations). The main findings were; the highest CU value was 86.7% in the square system of 2520 sprinkler under 200 kPa, 0.5 m height and 0.855 m3/h (Nozzle 2.5 mm). Meanwhile, in the triangular system, it was 87.3% under the same pressure and discharge and 1 m height. For applied machine learning, the highest values of R2 were 0.796, 0.825 and 0.929 in RF, XGB and XGB-RF respectively in the first scenario for CU. Moreover, for the DU, the highest values of R2 were 0.701, 0.479 and 0.826 in RF, XGB and XGB-RF respectively in the first scenario. The obtained results revealed that the sprinkler height had the lowest impact on modeling of the water distribution uniformity.