Artificial Neural Network Model Research Articles

A novel ANN-CA and MCDA integrated framework for predicting urban expansion and its implications on future flood risk, Accra Metropolis

Urban development in African countries significantly impacts environmental sustainability and city resilience, particularly in flood risk management. The Accra Metropolis, in particular, faces increasingly prevalent annual floods that disproportionately affect the urban poor living in informal settlements, resulting in significant loss of life and disruption of livelihoods. This study developed a GIS-based flood risk mapping framework that integrates artificial neural network (ANN) and cellular automata (CA) modelling with multi-criteria decision analysis. This hybrid approach was employed to predict flood scenarios for the Metropolis in 2032 and 2042. The framework incorporated six hydro-geomorphological indicators influencing extreme events: LULC, slope, elevation, drainage density, soil type, and proximity to rivers. The study analysed LULC projections, revealing a trend of substantial urban expansion with an estimated 10.9% increase in built-up areas within the next 20 years. This growth is expected to significantly heighten future flood risk in both the central upstream and downstream portions of the Metropolis, particularly in low-lying informal settlements close to the Odaw River and Korle Lagoon. Model performance was validated by ROC curve analysis, with AUC values of 0.927 and 0.922 for 2032 and 2042, respectively, which resonates with the records of previous flood distribution studies of the area. This study highlights the crucial need for sustainable urban development, improved infrastructure, and proactive flood mitigation measures in Accra to assist urban planners, policymakers, and stakeholders in managing flood risks effectively.

Integrating machine learning with experimental investigation for optimizing photocatalytic degradation of Rhodamine B using neodymium-doped titanium dioxide: a comprehensive approach with toxicity assessment.

In this study, neodymium-doped titanium dioxide (Nd-TiO2) nanoparticles were synthesized via a hydrothermal method for the photocatalytic degradation of Rhodamine B (RhB) under UV and sunlight conditions. The properties of these NPs were comprehensively characterized. And optimization of RhB degradation was conducted using control-variable experiment and artificial neural networks (ANN) under various operational conditions and in the presence of competing compounds. The acute toxicity of both NPs, RhB, and the environmental impact of the photocatalytic treatment effluent on Danio rerio were evaluated. The Nd modification increased the catalyst's specific surface area and thermal stability. X-ray diffraction confirmed the tetragonal anatase phase in undoped TiO2, while Nd-doped TiO2 exhibited shifts in peaks and the presence of brookite and rutile phases. Nd (1mol%) doped TiO2 demonstrated superior RhB photocatalytic degradation efficiency, achieving 95% degradation and 82% total organic carbon (TOC) removal within 60min under UV irradiation. Optimization under sunlight conditions yielded 95.14% RhB removal with 0.28g/L photocatalyst and 1% doping. Under UV light, 98.12% RhB removal was optimized with 0.97% doping, along with the presence of humic acid and CaCl2. ANN modeling achieved high precision (R2 of 0.99) in modeling environmental photocatalysis. Toxicity assessments indicated that the 96-h LC50 values were 681.59mg L-1 for both NPs, and 23.02mg L-1 for RhB. The treated dye solution exhibited a significant decline in toxicity, emphasizing the potential of 1% Nd-TiO2 in wastewater treatment.

Multi-Modal Machine Learning to Predict the Energy Discharge Levels from a Multi-Cell Mechanical Draft Cooling Tower

An artificial neural network was developed to augment the accuracy of a physically based computer model in relating heat discharge to visible plume volume of a 12-cell mechanical draft cooling tower. In a previous study, Savannah River National Laboratory developed a 1D model to capture the average power plant discharge levels via analysis of a series of visual images but was unable to accurately predict individual cases, resulting in an overall average error of about 5%, but individual comparisons resulted in an R2 of 0.36. Three optimization algorithms were applied to better fit the entrainment coefficients, and the artificial neural network model was applied to 289 cases of a 12-cell mechanical draft cooling tower power generation facility. Two artificial neural networks configurations consisted of 10 and 47 nodes that used as input readily available plant data, observed cooling tower plume conditions, observed operational conditions, local and regional weather, and the predicted plume volume from the physical model; the individual predictions’ accuracy improved to R2>0.95. This article concludes the sensitivities for the 1D model and additional actions to progress this field of study as well as applications for cooling tower monitoring. This strategy demonstrated an encouraging first step towards using multi-modal artificial neural network machine learning technology for information fusion to estimate power levels from external observations.

Superiority of artificial neural networks over conventional hydrological models in simulating urban catchment runoff

ABSTRACT The synergistic impacts of climate change and urbanisation have amplified the recurrence and austerity of intense rainfall events, exacerbating persistent flooding risk in urban environments. The intricate topography and inherent non-linearity of urban hydrological processes limit the predictive accuracy of conventional models, leading to significant discrepancies in flow estimation. Recent advancements in artificial neural network (ANNs) have demonstrated remarkable progress in mitigating most limitations, specifically in simulating complex, non-linear relationships, without an intricate comprehension of the underlying physical processes. This paper proposes a deep learning ANN-based flow estimation model for enhanced precision simulation of streamflow in urban catchments, with the research's distinctive contribution involving rigorous comparative evaluation of the developed model against the established Australian hydrological model, RORB. Gardiners Creek catchment, an urban catchment situated in East Melbourne was designated as the study area, with the model being calibrated upon historical storm incidences. The findings reveal that the ANN model substantially outperforms RORB, as evidenced by superior correlation, prediction efficiency, and lower generalisation error. This underscores the ANN's adeptness in accurately replicating non-linear-catchment responses to storm events, marking a substantial advancement over conventional modelling practices and indicating its transformative potential for enhancing flood prediction precision and revolutionising current estimation practices.

Evaluation of textile effluent treatment plant sludge as supplementary cementitious material in concrete using experimental and machine learning approaches

The growing demand for readymade garments is leading to the production of a large amount of textile effluent sludge (TES) from wastewater treatment plants, which is currently disposed of as backfill, causing land and groundwater contamination. This scenario has prompted researchers to explore the use of industrial waste, such as TES, as a supplementary cementitious material (SCM) to mitigate the carbon footprint of cement production. Therefore, this study aims to investigate the fresh, mechanical, non-destructive characteristics, and predictive modeling of mechanical strengths of TES-based concrete. Machine learning models, including random forest (RF) and artificial neural networks (ANN), were utilized to predict compressive and tensile strengths from various input variables. This study also evaluated the alkalinity, specific gravity, oxide content, and surface area of TES. The findings showed a pH of 9.4 and a CaO content of 33.3 %, indicating TES's potential as an SCM. The research further investigated the properties of concrete with 5–20 % cement-replaced TES, using a total binder content of 488.1 kg/m3 and a water-binder ratio of 0.32. The fresh state properties of the 5–15 % TES incorporated concrete produced normal density concrete, with a slump of 100 ± 20 mm, Kelly ball penetration of 45 ± 7 mm, a compaction factor of 0.92 ± 0.05, and a fresh density of 2464 ± 152 kg/m3. At 28 days, concrete with 15 % TES replacement showed a compressive strength of 25.6 MPa and a tensile strength of 2.88 MPa, achieving 76 % of the strength activity index and 72 % of the relative tensile strength. However, the 15 % TES concrete displayed a water permeability coefficient of 4.4 × 10−12 m/s and an electrical resistivity of 31.5 kΩ cm, compared to the control mix's 2.0 × 10−12 m/s and 37.7 kΩ cm. However, the hardened properties declined significantly after replacing 15 % of cement with TES. The mechanical properties were also estimated using ultra pulse velocity and rebound hammer tests, showing a coefficient of determination (R2) higher than 0.87. The RF model demonstrated superior performance with an R2 value of 0.8802 for compressive strength and 0.8518 for tensile strength, outperforming the ANN model's R2 values of 0.8126 and 0.695, respectively. Scanning electron microscopy confirmed the consistent development of hydration products in the TES-enhanced concrete samples.

Prediction of Total Soluble Solids Content Using Tomato Characteristics: Comparison Artificial Neural Network vs. Multiple Linear Regression

Tomatoes are among the world’s most significant vegetables, both in terms of production and consumption. Harvesting takes place in tomato production when the important quality attribute of total soluble solids content reaches its maximum possible level. Tomato total soluble solids content (TSS) is among the most crucial attribute parameters for assessing tomato quality and for tomato commercialization. Determination of total soluble solids content by conventional measurement methods is both destructive and time-consuming. Therefore, the tomato processing industry needs a rapid identification method to measure total soluble solids content (TSS). In this study, we aimed to estimate how much soluble solids there are in beef tomato fruit by Artificial Neural Networks (ANN) and Multiple Linear Regression (MLR) methods. The models were assessed using the Coefficient of Determination (R2), Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE) metrics. The training data set results of the MLR model established to estimate the amount of brix in tomato fruit, calculated as MAE: 0.2349, RMSE: 0.3048, R2: 0.8441, and MAPE: 5.5368, while, according to the ANN model, MAE: 0.0250, RMSE: 0.031, R2: 0.9982 and MAPE: 0.5814. According to the metric outcomes, the ANN-based model performed better in both the training and testing parts.

Artificial Intelligence Using FFNN Models for Computing Soil Complex Permittivity and Diesel Pollution Content

Soil pollution caused by hydrocarbons, such as diesel, poses significant risks to both human health and the ecosystem. The evaluation of soil pollution and various soil engineering applications often relies on the analysis of complex permittivity, encompassing parameters such as dielectric constant and dielectric loss. Various computational models, including theoretical physics-based models, mixture theory models, statistical empirical models, and artificial neural network (ANN) models, have been explored for computing soil complex permittivity and predicting water and pollutant content. Theoretical models require detailed data that is often unavailable, and thus have limited applicability. Mixture models tend to underestimate soil characteristics due to inaccuracies in permittivity estimation of soil phases. While empirical models are widely used, their applicability is restricted to specific soil types, datasets, and locations. ANN models offer promising predictions, accommodating nonlinear phenomena and allowing for missing information and variables. In this study, capacitive electromagnetic electrode sensors were utilized to determine the complex permittivity of soil contaminated with varying levels of diesel at different moisture levels. Theoretical mixture, empirical, and Feed Forward Neural Network (FFNN) models were employed to compute the permittivity of polluted soil based on its phases and to predict the level of diesel pollution. A comparison of these modeling approaches revealed that the FFNN model exhibited the best performance. The ANN model demonstrated superior performance metrics, including a high correlation coefficient and lower mean square error. Specifically, the correlation coefficients for the FFNN model were 0.9942 for training samples, 0.9967 for validation samples, and 0.9977 for test samples. Additionally, the ANN model yielded the lowest mean square error compared to the other three models. Doi: 10.28991/CEJ-2024-010-09-018 Full Text: PDF

Predictive modeling of MRR, TWR, and SR in spark-EDM of Al-4.5Cu–SiC using ANN and GEP

In this study, Al-4.5Cu alloy was reinforced with varying weight percentages of SiC particles (2%, 4%, 6%, and 8%) to create metal matrix composites via the stir casting method. The formation of intermetallic compounds was confirmed through energy dispersive spectroscopy and x-ray diffraction analysis. This article compares the performance of Artificial Neural Network (ANN) and Gene Expression Programming (GEP) models in predicting the Metal Removal Rate (MRR), tool wear rate, and surface roughness in the die-sinking electro-discharge machining (EDM) process of the ex-situ developed Al-4.5%Cu–SiC composites. The study considers three machine parameters—pulse on time (TON), pulse off time (TOFF), and current (I)—along with the weight fraction of SiC particles as input variables for the models. Both ANN and GEP models demonstrated high predictive accuracy for the EDM performance metrics, with correlation coefficients (R) ranging from 0.973 68 to 0.980 65 for the ANN model and 0.980 11 to 0.982 59 for the GEP model. Notably, the GEP model exhibited superior predictive capability, as evidenced by its higher correlation coefficients and lower root mean square error, indicating greater effectiveness in predicting the EDM process outcomes than the ANN model.

A comprehensive analysis of the artificial neural networks model for predicting monkeypox outbreaks

Monkeypox is a viral disease that causes outbreaks in various countries, significantly impacting public health and healthcare systems. Effective preparedness and response efforts require accurately predicting the severity of these outbreaks. Currently, there are no publicly released studies for nations like Chile and Mexico on monkeypox, leading to this study's creation. We use a neural network model with a time series dataset of monkeypox cases from multiple countries, including Argentina, Brazil, France, Germany, Chile, and Mexico. The Levenberg-Marquardt learning technique is employed to develop and validate single and two hidden layers artificial neural network models. We train various model architectures with different numbers of hidden layer neurons using the K-fold cross-validation early stopping method. Additionally, we use long short-term memory and gated recurrent unit models, commonly employed for time series data processing, to compare the performance of our artificial neural network model.

Artificial neural network-based models for short term forecasting of solar PV power output and battery state of charge of solar electric vehicle charging station

The main objective of this study is to develop ANN-based predictive models for short-term forecasting of solar PV power output and battery state of charge. The 3Ds energy model that integrates Decarbonization, Digitalization, and Decentralization of the energy system to facilitate the shift towards sustainable energy sources is used for this electric vehicle project. The experimental set up includes solar PV panels, a solar inverter, a battery inverter, and a battery bank. Additionally, smart meters were installed to collect real-time performance data from the solar PV-powered electric vehicle (EV) charging station. The weather and real-time system performance data were used to develop short term forecasting models based on Artificial Neural Networks (ANN) and the Levenberg-Marquardt method, to predict the performance of the system (power output and state of the charge) ahead. The R values for the prediction models of solar photovoltaic (PV) specific power and battery state of charge fall within the range of 0.9957–0.9969 and 0.9990 to 0.9996, respectively. Furthermore, the mean squared errors (MSEs) for the artificial neural network (ANN) models pertaining to solar photovoltaic (PV) power output and battery state of charge exhibit a variety of values. Specifically, the MSEs for solar PV power output range from 1.242 x 10^-4 to 1.579 x 10^-4, while the MSEs for battery status of charge range from 0.1889 to 0.3402. Artificial neural network (ANN) models possess considerable promise for practical implementations as they simplify intricate connections among inputs, parameters, and outputs in real-world scenarios. The level of accuracy and short-term predictive models of the solar PV powered EV charging station are very important for achieving a balance between the supply of solar photovoltaic (PV) system and the demand for electric vehicles (EVs). Predictive models will help build complex control mechanisms for controlling and optimizing solar PV-powered charging stations, supervising their operation and maintenance, and simplifying renewable energy pre-purchase. Future works will include the development of an energy management system to improve the efficiency of the solar stations charging the EVs, the establishment of blockchain networks for both the solar PV system and battery bank, and the integration of cybersecurity protocols for the charging station.

Data-Driven Multi-Fault Detection in Pipelines Utilizing Frequency Response Function and Artificial Neural Networks

This research presents a data-driven structural health monitoring (SHM) approach for pipeline systems that leverages frequency response function (FRF) signals and artificial neural network (ANN) algorithms to accurately identify and classify diverse pipeline fault conditions. The study focuses on three specific faults: bolt looseness, scale deposits, and crack occurrence at pipeline supports, which were replicated on a pipeline segment located at the Sound and Vibration Research Group (SVRG) at University Putra Malaysia (UPM). The FRF signals were captured using accelerometers to monitor the structural health of the pipeline. The data acquisition stage involved collecting FRF signals from the accelerometers to capture vibrations and responses related to the identified faults using a Siemens LMS SCADAS data acquisition unit. The data underwent preprocessing, including the application of principal component analysis (PCA) for feature selection. The subsequent data processing stage involved the application of an artificial neural network (ANN) algorithm for pattern recognition to analyze and classify the acquired data, identifying patterns associated with the replicated fault conditions. The proposed methodology demonstrated exceptional performance, with the ANN model achieving consistently high overall accuracy (above 99.7%) and remarkably low mean squared error (in the range of 0.0088×10−3to0.3062×10−3) across multiple iterations and sensor datasets. The detailed class-specific metrics, including accuracy, precision, sensitivity, and F1-score, further substantiated the model's effectiveness in identifying the individual fault types with near-perfect or perfect results for the majority of the fault scenarios. The location-invariant performance of the ANN model across different sensor placements demonstrates the robustness of the proposed data-driven SHM methodology. This research highlights the transformative potential of integrating state-of-the-art data-driven techniques to revolutionize the monitoring and assessment of critical pipeline infrastructure, ultimately enhancing the safety, reliability, and longevity of these vital systems.

Development of machine learning-based standalone GUI application for predicting hydraulic conductivity and compaction parameters of lateritic soils

Hydraulic conductivity and compaction parameters are the key considerations in selecting lateritic soils for engineering construction. Nevertheless, the complexity and high cost of the required tests have driven many contractors and field engineers to skip them, resulting in a succession of engineering structure failures. To overcome this limitation, this study developed machine learning-based standalone GUI application to predict lateritic soils’ hydraulic conductivity (K), maximum dry density (MDD) and optimum moisture content (OMC) from indices including specific gravity, liquid limit, plasticity index, linear shrinkage and fine content. To achieve this goal, the geotechnical properties of three hundred samples, collected using grid sampling method from thirty different lateritic deposits in southwestern Nigeria, were evaluated through laboratory tests. The test results were used to train predictive models using artificial neural network (ANN), adaptive neuro-fuzzy inference system (ANFIS) and Gaussian process regression (GPR). The models’ performance was compared using coefficient of determination (R2), root mean squared error (RMSE), mean absolute percentage error (MAPE) and mean absolute error (MAE). Based on these performance metrics, ANN demonstrated the best performance (R2 = 0.9835, 0.9797, 0.9999; RMSE = 7.938, 0.252, 2.09E-08; MAPE = 0.288, 1.114, 1.587; MAE = 5.432, 0.169, 1.1E-08) for MDD, OMC and K, respectively, followed by GPR and then ANFIS. Thus, the ANN models were selected and embedded in a standalone GUI application to enhance easy and quick prediction of lateritic soils’ MDD, OMC and K. The validity of the ANN-based standalone GUI application was demonstrated by comparing it favorably to notable regression-based models in the literature.

Machine learning assisted mechanism modeling for gas phase electrohydrodynamic system

In this paper, a hybrid physics-data driven model for electrohydrodynamic gas system (EHDGS) was developed by combining artificial neural network (ANN) with mechanism modeling method. ANN was used to correlate the relationship between the variables (electrode distance, diameter of grounding cylinder, applied voltage, electric field gradient, etc.) in a needle-cylinder EHDGS and the initial space charge density. The results showed that the ANN model of nine neurons can well predict the initial space charge density. The coefficient of determination (R2) reaches 0.9874, and the mean absolute error is as low as 0.0067. Subsequently, a hybrid mechanism model where the initial space charge density was predicted from the ANN model was constructed to simulate the needle-cylinder EHDGS. The experiment with the needle-cylinder EHDGS was carried out. The simulation results were in good agreement with the experimental data, demonstrating the reliability of the proposed hybrid model. The electric field distribution, space charge distribution, and flow field distribution behavior of the EHDGS were then analyzed in detail. The effects of key parameters on the flow characteristics of EHDGS were systematically studied, showing that higher voltage and shorter distance give higher flow rate up to 2.5 m/s. The diameter of the cylinder also significantly influences the breakdown voltage. Three dimensionless groups were defined and their effects on spatial charge density distribution were investigated. This study provides both insights and an efficient tool for the design and optimization of EHDGS.

Assessing the performance of a monocrystalline solar panel under different tropical climatic conditions in Cameroon using artificial neural network

To implement the European Union (EU)-Africa Green Energy Initiative in Cameroon to boost the renewable energy sector, we model the performance of a 500 W monocrystalline solar panel in major cities of Cameroon located in different climatic zones to select the best location for the installation of a solar farm. We also evaluate the contribution of seasonal and weather variability to the amount and stability of power generated by the panel using the artificial neural network (ANN). The ANN model was used to train and test the ERA5 hourly data for Bamenda. The model was then used to estimate Photovoltaic (PV) output in Douala, Yaounde, Ngaoundere, Garoua, and Maroua with a mean absolute error of 4.109 × 10−5, 4.699 × 10−5, 3.563 × 10−5, 3.106 × 10−5, and 3.083 × 10−5 kW, respectively. The results show that the ANN can capture the influence of weather variability on the generated output power. Cloud cover and rainfall are found to negatively affect the amount and stability of generated power in the lower latitude cities of Douala and Yaounde compared to the northern cities, with these effects being stronger in the rainy season than in the dry season. Garoua followed by Maroua are proving to be the best locations for installing a solar park in terms of the amount and stability of electricity generated throughout the year. The Cameroonian government, its EU partners, and other stakeholders involved in the development of solar energy in the country will be able to use the results of this study for better decision-making.

Artificial neural network enabled photovoltaic-thermoelectric generator modelling and analysis

Photovoltaic-Thermoelectric Generator (PV-TEG) system has emerged as a promising approach to significantly enhance the efficiency of conventional PV cells. However, optimizing the performance of these hybrid systems presents a formidable challenge due to their complex structure and multitude of design parameters. This study tackles such challenge by developing a machine learning based Artificial Neural Network (ANN) model which comprises two sub-ANN models that can work independently for PV and TEG modules or in combination through a cyclic approach for the hybrid PV-TEG system. The model demonstrates remarkable versatility, allowing control over various parameters such as PV coating, device geometry, and environmental conditions. Compared to COMSOL simulations, the ANN model achieves over 97.6 % accuracy with a 6000-fold increase in simulation speed, enabling extensive parameter sweeps and insightful system analysis. Within 18 min, the model conducted a real-time simulation using 8712 weather data entries from Singapore in 2022 and predicted that the hybrid PV-TEG system would generate a total power of 265 kWh/m2, 6.4 % more than that of the standalone PV system with an average system temperature reduction of 7 K. The model's rapid processing capabilities and high accuracy are particularly beneficial for large-scale simulations and practical applications in renewable energy technology.

Enhanced heat transfer in novel star-shaped enclosure with hybrid nanofluids: A neural network-assisted study

The current research is a numerical investigation of the thermo-fluidic transport trend within a new star-shaped enclosure filled with Fe3O4−MWCNT (Multiwall Carbon Nanotubes) suspended hybrid nanoparticles in a base fluid (water), induced by a horizontal magnetizing field. A rectangular rod inside the cavity positioned at center is kept at a high temperature, as the low temperature is chosen for outer corrugated boundary. Formulation yelled in mathematical principles, along with the boundary conditions and some physical assumptions, are formulated as dimensionless PDEs. The finite element method (FEM) assisted by “COMSOL” software, employed for numerical simulations, and the PARDISO solver is used for solving nonlinear system of equations. Also, an artificial neural network (ANN) model is assembled to analyze the effect of these parameters on thermal and flow transport properties of Hybrid nanofluid. The training of this ANN model is done by a dataset generated from numerical simulations, suggesting predictions regarding heat transport under varying parameters. This approach does not only reduce the effort, but also reduces computing time when exploring the thermal behaviors across various datasets. The visualization of velocity and temperature fields through streamlines and isotherms, along with the evaluation of the Nusselt number, illustrate the enhanced heat transfer facilitated by the incorporation of Fe3O4−MWCNT nanoparticles. The results show that increased magnetic field strength reduces fluid velocity due to the generated Lorentz force, which counteracts convective flow. Percentage analysis indicates that hybrid nanofluids (Fe3O4−MWCNT) significantly improve thermal distribution compared to conventional fluids. This study demonstrates the effectiveness of ANN models in forecasting the influence of diverse factors on the flow and heat transport characteristics of hybrid nanofluids. These insights are essential to the design and optimization of heat transfer systems.

Predicting pavement surface conditions through artificial neural networks

In this study, six distinct types of pavement distress were evaluated using the pavement condition index (PCI) rating: ravelling, rutting, cracking (including alligator, transverse and longitudinal cracks), shoving, patching and potholes. The severity levels of these distress types are utilised for classification, and these levels are subsequently converted into PCI values. The proposed methodology incorporates a thorough analysis, considering various variables such as average daily traffic, land-use classification, number of lanes, road width and length, pavement age, road alignment, vehicle type and weather conditions in Berzhite, Tirana, Albania. To predict PCI values, an artificial neural network model was employed due to its versatility in handling diverse inputs. Key performance metrics, including R2, mean squared error and mean absolute error, are used for model assessment. Notably, model 2, featuring two hidden layers, exhibits superior performance over model 1. However, due to size constraints in testing and validation datasets in this study, the accuracy of the model is somewhat limited. Future directions for research involve expanding the dataset to enhance model accuracy and incorporating advanced features to capture a more complex comprehension of pavement conditions.

Machine learning approaches for assessing stability in acid-crude oil emulsions: application to mitigate formation damage

The stability of acid-crude oil emulsion poses manifold issues in the oil industry. Experimentally evaluating this phenomenon may be costly and time-consuming. In contrast, machine learning models have proven effective in predicting and evaluating various phenomena. This research is the first of its kind to assess the stability of acid-crude oil emulsion, employing various classification machine learning models. For this purpose, a data set consisting of 249 experimental data points belonging to 11 different crude oil samples was collected. Three tree-based models, namely decision tree (DT), random forest (RF), and categorical boosting (CatBoost), as well as three artificial neural network models, namely radial basis function (RBF), multi-layer perceptron (MLP) and convolutional neural network (CNN), were developed based on the properties of crude oil, acid, and protective additive. The CatBoost model obtained the highest accuracy with 0.9687, followed closely by the CNN model with 0.9673. In addition, confusion matrix findings showed the superiority of the CatBoost model. Finally, by applying the SHapley Additive exPlanations (SHAP) method to analyze the impact of input parameters, it was found that the crude oil viscosity has the most significant effect on the model's output with the mean absolute SHAP value of 0.88.

Zinc (II)-1,4,7,10-tetraazacyclododecane complexes catalyzed tertiary amine-based CO2 capture in a rotating packed bed

A novel tertiary amine-based CO2 absorption process using zinc (II)-1,4,7,10-tetraazacyclododecane complexes (CN) as a catalyst in a rotating packed bed was proposed for energy-efficient CO2 capture. The CN-catalysis reaction mechanism was first studied by density functional theory calculation, and the Gibbs Free Energy profile was established. Furthermore, the effects of various operation conditions for the tertiary amine-CN-CO2 systems on the CO2 removal efficiency (η), the overall gas-phase volumetric mass transfer coefficient (KGa), and the efficiency enhancement factor (Φ) were investigated. It was found that η and KGa with adding 4 wt% CN were 1–3 folds than those uncatalyzed. The absorption performance of CO2 in DEEA solution with CN was superior to that in 10 wt% DEA solution. Finally, an artificial neural network model was established to predict KGa, and the predicted data agreed well with experimental results with the average absolute relative deviation of 5.07 %.

Development and Validation of Upper Limb Lymphedema in Patients After Breast Cancer Surgery Using a Practicable Machine Learning Model: A Retrospective Cohort Study.

Upper limb lymphedema is one of the most common adverse events related to surgery owing to the large gap between guideline implementation and the intended clinical outcomes. However, the monitoring of limb lymphedema remains challenging because of vague clinical presentations. This study aimed to develop and validate practical predictive models for upper limb lymphedema through machine learning. We retrospectively collected clinical data to develop models for early risk prediction of upper limb lymphedema based on a single-center electronic health record data from patients who underwent breast cancer surgery from June 2021 through June 2023. For prediction model building, 70% and 30% of the data were randomly split into training and testing sets, respectively. We then developed an upper limb lymphedema prediction model using machine learning algorithms, which included random forest model (RFM), generalized logistic regression model (GLRM), and artificial neural network model (ANNM). For evaluating the model's performance, we used the area under the receiver operating characteristic curve (AUROC), calibration curve to compare different models. The potential clinical usefulness of the best model at the best threshold was assessed through a net benefit approach using a decision curve analysis (DCA). Of the 3201 patients screened for eligibility, 3160 participants were recruited for the prediction model. Among these, Body Mass Index (BMI), hypertension, TNM, lesion site, level of lymph node dissection(LNMD), treatment, and nurse were independent risk factors for upper limb lymphedema and were listed as candidate variables of ML-based prediction models. The RFM algorithm, in combination with seven candidate variables, demonstrated the highest prediction efficiency in both the training and internal verification sets, with an area under the curve (AUC) of 0.894 and 0.889 and a 95% confidence interval (CI) of 0.839-0.949 and 0.834-0.944, respectively. The other two types of prediction models had prediction efficiencies between AUCs of 0.731 and 0.819 and 95% CIs of 0.674-0.789 and 0.762-0.876, respectively. The interpretable predictive model helps physicians more accurately predict the upper limb lymphedema risk in patients undergoing breast cancer surgery. Especially for the RFM, this newly established machine learning-based model has shown good predictive ability for distinguishing high risk of upper limb lymphedema, which could facilitate future clinical decisions, hospital management, and improve outcomes.