Group Method Of Data Handling Research Articles

Introducing effective parameters for predicting job burnout using a self-organizing method based on group method of data handling neural network.

In addition to affecting people's bodily and mental health, the Covid-19 epidemic has also altered the emotional and mental well-being of many workers. Especially in the realm of institutions and privately held enterprises, which encountered a plethora of constraints due to the peculiar circumstances of the epidemic. It was thus anticipated that the present study would use a group method of data handling (GMDH) neural network for analyzing the relationship of demographic factors, Coronavirus, resilience, and the burnout in startups. The test methodology was quantitative. The research examined 384 startup directors and representatives, which is a sizable proportion of the limitless community. The BRCS, the MBI-GS, and custom-made assessments of stress due to the Coronavirus were all used to collect data. Cronbach's alpha confirmed the polls' dependability, and an expert panel confirmed the surveys' authenticity. The GMDH neural network's inherent potential for self-organization was used to choose the most useful properties automatically. The trained network has a three-layered topology with 4, 3, and 2 neurons in each of the hidden layers. The GMDH network has significantly reduced the computational load by using just 7 parameters of marital status, stress of covid-19, job experience, professional efficiency, gender, age, and resilience for burnout categorization. After comparing the neural network's output with the acquired data, it was determined that the constructed network accurately classified all of the information. Among the achievements of this research, high accuracy in predicting job burnout, checking the performance of neural network in determining job burnout and introducing effective characteristics in determination of this parameter can be mentioned.

A novel neural-evolutionary framework for predicting weight on the bit in drilling operations

This study compares the performance of artificial neural networks (ANN) trained by grey wolf optimization (GWO), biogeography-based optimization (BBO), and Levenberg–Marquardt (LM) to estimate the weight on bit (WOB). To this end, a dataset consisting of drilling depth, drill string rotational speed, rate of penetration, and volumetric flow rate as input variables and the WOB as a response is used to develop and validate the intelligent tools. The relevance test is applied to sort the strength of WOB dependency on the considered features. It was observed that the WOB has the highest linear correlation with the drilling depth and drill string rotational speed. After dividing the databank into the training and testing (4:1) parts, the proposed LM-ANN, GWO-ANN, and BBO-ANN ensembles are constructed. A sensitivity analysis is then carried out to find the most powerful structure of the models. Each model performs to reveal the relationship between the WOB and the mentioned independent factors. The performance of the models is finally evaluated by mean square error (MSE) and mean absolute error criteria. The results showed that both GWO and BBO algorithms effectively help the ANN to achieve a more accurate prediction of the WOB. Accordingly, the training MSEs decreased by 14.62% and 24.90%, respectively, by applying the GWO and BBO evolutionary algorithms. Meanwhile, these values were obtained as around 9.86% and 9.41% for the prediction error of the ANN in the testing phase. It was also deduced that the BBO performs more efficiently than the other technique. The effect of input variables dimension on the accuracy and training time of the BBO-ANN clarified that the most accurate WOB predictions are achieved when the model constructs with all four input variables instead of utilizing either three or two of them with the highest linear correlation. It was also observed that the training stage of the BBO-ANN model with four input variables needs a little more computational time than its training with either two or three variables. Finally, the accuracy of the BBO-ANN model for the WOB prediction has been compared with the multiple linear regression, support vector regression, adaptive neuro-fuzzy inference systems, and group method of data handling. The statistical accuracy analysis confirmed that the BBO-ANN is more accurate than the other checked techniques.

EPNN-based prediction of meteorological data for renewable energy systems

In this paper, an application of an Evolving Polynomial Neural Network (EPNN) for prediction of meteorological time series (global solar irradiation, air temperature, relative humidity, and wind speed) is described. Prediction of such data plays a very important role in design of the renewable energy systems. The problem of time series prediction is formulated as a system identification problem, where the input of the system is the past values (y (t - 1), y (t - 2), y (t - 3), …) of a time series and its desired output (y (t), y (t + 1), y (t + 2), …) are the future of a time series. In this study, a dataset of meteorological time series for five years collected in Algiers (Algeria) by the National Office of Meteorology has been used. The obtained results showed a good agreement between both series, measured and predicted. The correlation coefficient ( r ) is arranged between 0.9821 and 0.9923, the mean relative error over the whole data set is not exceed 15.4 %. The proposed model provides more accurate results than other ANN’s architecture, wavenet (wavelet-network) and Adaptive Neuro-Fuzzy Inference Scheme (ANFIS). In order to show the effectiveness of the proposed predictor, the predicted data have been used for sizing, and prediction of the output energy of photovoltaic systems.

Selection of effective combination of time and frequency features using PSO-based technique for monitoring oil pipelines

Pipeline installation is a time-consuming and expensive process in the oil sector. Because of this, a pipe is often utilized to carry diverse petroleum products; hence, it is crucial to use a precise and dependable control system to identify the kind and quantity of oil products being transported. This study attempts to identify four petroleum products by using an X-ray tube-based system, feature extraction in the frequency and temporal domains, and feature selection using Particle Swarm Optimization (PSO) in conjunction with a Group Method of Data Handling (GMDH) neural network. A sodium iodide detector, a test pipe that simulates petroleum compounds, and an X-ray source make up the implemented system. The detector's output signals were transmitted to the frequency domain, where the amplitudes of the top five dominant frequencies could be determined. Furthermore, the received signals were analyzed to extract five temporal characteristics-MSR, 4th order moment, skewness, WL, and kurtosis. The PSO system takes into account the extracted time and frequency features as input in order to introduce the optimal combination. Four different GMDH neural networks were constructed, and the chosen characteristics were used as inputs for those networks. Finding the volume ratio of each product was the responsibility of each neural network. The four designed neural networks were able to predict the amount of ethylene glycol, crude oil, gasoil, and gasoline with RMSE of 0.26, 0.17, 0.19, and 0.23, respectively. One compelling argument for using the proposed approach in the oil industry is that it can calculate the volume ratio of products with a root mean square error of no more than 0.26. The adoption of a feature selection method to choose the best ones is credited with this remarkable degree of precision. By providing appropriate inputs to neural networks, the control system has significantly outperformed its predecessors in terms of precision and efficiency.

Chebyshev polynimal dempster convolutional hearing sensitivity level detection with auditory evoked potentials

The EEG-based HTR utilizing AEP responses of both group of participants with normal hearing and abnormal hearing are managed with the objective of detecting hearing sensitivity level using Chebyshev Recurrence Polynomial and Dempster Convolutional Neural Network (CRP-DCNN) is designed. The CRP-DCNN method is split into three sections. They are preprocessing using Chebyshev Recurrence Polynomial Filter, feature extraction by employing Orthogonalized Singular Value and Median Skewed Wavelet. Here, both Orthogonalized Singular Value Decomposition-based parametric and Median Skewness-based non-parametric modeling techniques are employed for first obtaining the hearing threshold factors and then extracting statistical features for further processing. Finally Dempster Convolutional Neural Network-based Classification for detecting hearing sensitivity level is presented. Hence, the objective to determine the significant correlations between the brain dynamics and the auditory responses and detect the hearing sensitivity level of the group of participants with normal hearing and with the group of participants with hearing loss are designed on accordance with the features of EEG signals. Simulations are performed in MATLAB to validate the features of EEG signals.

Machine-learning-based short-term forecasting of daily precipitation in different climate regions across the contiguous United States

Forecasting the precipitation is needed for water resources management and planning, irrigation scheduling and flood modeling. The purpose of this research is to obtain short-term forecasts of the daily precipitation (1-, 2-, and 3-day-ahead) at 28 sites located in various climate regions over the contiguous United States. To this end, daily precipitation data for the study sites are collected from 1995 to 2019. Data from 1995 to 2014 and from 2015 to 2019 are considered respectively in the training and forecasting steps. The precipitation forecasts are obtained using three artificial intelligence models: wavelet particle swarm optimization adaptive neuro-fuzzy inference system (WPSOANFIS), wavelet group method of data handling (WGMDH), and wavelet long short-term memory (WLSTM). The 28-site-average mean absolute errors of 1-day-ahead precipitation forecasts from WLSTM, WGMDH, and WPSOANFIS are respectively, 0.65, 0.77, and 1.31 mm/d. The 28-site mean root mean square error (RMSE) and coefficient of determination (R2) for the WLSTM are 1.47 mm/d and 0.91, respectively. The average RMSE and R2 for WGMDH (WPSOANFIS) are respectively 1.66 mm/d and 0.88 (3.09 mm/d and 0.59). WGMDH outperforms WLSTM and WPSOANFIS in 2- and 3-day-ahead precipitation forecasting. The machine learning techniques show their best results in the West and Southwest climate zones that have the lowest standard deviation of daily precipitation measurements. Lower accuracies are observed for sites with the largest standard deviations of daily precipitation observations, particularly in the 2- and 3-day-ahead horizons. Deep learning techniques (i.e., WGMDH and WLSTM) generate more accurate precipitation forecasts than WPSOANFIS. The 1-, 2-, and 3-day-ahead precipitations obtained using the proposed models are more accurate than those obtained using existing models.

An Optimal Operation Strategy for Collaborative Flexibility Provision of a Carbon Capture and Utilization Process With Wind Energy

Improving power system flexibility by responsive demand is essential for integrating wind energy with a high level of variability in power systems. Carbon dioxide-based chemical processes as energy-intensive industrial loads may offer a vast potential of new forms of flexible operation due to their existing control infrastructure and storage capabilities. However, a collaborative decision model is needed for optimal energy sharing among the chemical plant and the grid under the variations and uncertainties of wind power. This study develops an optimal two-stage stochastic programming model for a novel flexible operation strategy of the chemical process coupled with wind turbines. In the proposed control scheme, a small-scale wind farm provides the power input of a chemical plant. Wind turbines are connected to the grid and actively participate in the day-ahead energy and reserve markets, considering the chemical plant as a source of flexibility. An equivalent scenario-based model of the proposed optimization problem is suggested using the Group Method of Data Handling (GMDH) for a data-driven prediction of stochastic variables. Simulation results demonstrate the effectiveness and significance of the proposed approach for an optimal and collaborative contribution in ancillary market of a carbon dioxide-based chemical plant supplied by wind energy.

Thermal efficiency of microchannel heat sink: Incorporating nano-enhanced phase change materials and porous foam gradient and artificial intelligence-based prediction

This study investigates the impact of incorporating phase change materials (PCMs), nano-particles (NPs) enhanced PCMs (nePCMs), and porous foam gradients on the thermal performance (TPEF) of microchannel heat sinks (MCHS). Specifically, the effect of different PCM types, hybrid NPs, and spiral microchannels on the TPEF is examined using numerical solutions based on the finite volume method. The results indicate that increasing Re and using spiral microchannels significantly enhance the TPEF. The incorporation of NPs-water and PCM can reduce the thermal resistance (R) of MCHS, with PCM significantly improving the TPEF. Among the PCMs, ENCAPSUL demonstrates the best performance for MCHS. The combination of hybrid nePCM increases TPEF by approximately 9%. The combination of PCM-aluminum oxide-iron oxide NPs exhibits the highest TPFE. The use of a porous medium with PCM can decrease the R by about 60 %, and it improves the TPEF by altering the conduction and convection mechanism. various scenarios of changes in the porosity coefficient have been considered for porous foam gradient and the best performance is achieved for the NYPC mode. Various scenarios of MCHS with PCM-water combinations have been explored, and the best performance is observed when PCM is situated in the microchannel. Additionally, artificial intelligence techniques, such as the Group Method of Data Handling (GMDH), have been utilized to estimate R, and a multivariate polynomial regression (MPR) equation has been developed to calculate R based on input variables. GMDH is more accurate compared to MPR.

A Deep GMDH Neural-Network-Based Robust Fault Detection Method for Active Distribution Networks

The increasing penetration of distributed generation (DG) to power distribution networks mainly induces weaknesses in the sensitivity and selectivity of protection systems. In this manner, conventional protection systems often fail to protect active distribution networks (ADN) in the case of short-circuit faults. To overcome these challenges, the accurate detection of faults in a reasonable fraction of time appears as a critical issue in distribution networks. Machine learning techniques are capable of generating efficient analytical expressions that can be strong candidates in terms of reliable and robust fault detection for several operating scenarios of ADNs. This paper proposes a deep group method of data handling (GMDH) neural network based on a non-pilot protection method for the protection of an ADN. The developed method is independent of the DG capacity and achieves accurate fault detection under load variations, disturbances, and different high-impedance faults (HIFs). To verify the improvements, a test system based on a real distribution network that includes three generators with a capacity of 6 MW is utilized. The extensive simulations of the power network are performed using DIgSILENT Power Factory and MATLAB software. The obtained results reveal that a mean absolute percentage error (MAPE) of 3.51% for the GMDH-network-based protection system is accomplished thanks to formulation via optimized algorithms, without requiring the utilization of any feature selection techniques. The proposed method has a high-speed operation of around 20 ms for the detection of faults, while the conventional OC relay performance is in the blinding mode in the worst situations for faults with HIFs.

Applying feature selection and machine learning techniques to estimate the biomass higher heating value

The biomass higher heating value (HHV) is an important thermal property that determines the amount of recoverable energy from agriculture byproducts. Precise laboratory measurement or accurate prediction of the HHV is essential for designing biomass conversion equipment. The current study combines feature selection scenarios and machine learning tools to establish a general model for estimating biomass HHV. Multiple linear regression and Pearson’s correlation coefficients justified that volatile matter, nitrogen, and oxygen content of biomass samples have a slight effect on the HHV and it is better to ignore them during the HHV modeling. Then, the prediction performance of random forest, multilayer and cascade feedforward neural networks, group method of data handling, and least-squares support vector regressor are compared to determine the intelligent estimator with the highest accuracy toward biomass HHV prediction. The ranking test shows that the multilayer perceptron neural network better predicts the HHV of 532 biomass samples than the other intelligent models. This model presents the outstanding absolute average relative error of 2.75% and 3.12% and regression coefficients of 0.9500 and 0.9418 in the learning and testing stages. The model performance is also superior to a recurrent neural network which was recently developed in the literature using the same databank.

NGHIÊN CỨU ỨNG DỤNG KỸ THUẬT TRÍ TUỆ NHÂN TẠO TRONG BÀI TOÁN DỰ BÁO GIÁ MỘT SỐ MẶT HÀNG

The global economy is significantly impacted by changes in the price of primary commodities. As a result, both the academic and professional sectors have paid attention to price predictions for major commodities. The goal of this study is to build an artificial intelligence-based model for one-day market price predictions for important commodities like copper, crude oil, gas, and silver. The information on commodity trading was gathered between 01/2000 and 10/2019. Different models based on group method of data handling (GMDH), long short-term memory (LSTM), artificial neural network (ANN), and adaptive neuro fuzzy inference system (ANFIS) were developed. Theil's U, RMSE, MAPE, MAE, R, and other performance indices were used to compare the models. The findings demonstrated that, in terms of commodity price prediction, the suggested model based on GMDH technique performs better than alternative approaches. A viable alternative for price prediction is the GMDH-based model. For economists and professionals involved in commodity price forecasting, the GMDH can be a useful tool.

Prediction of the thermal conductivity of rocks using group method of data handling (GMDH)

The effective thermal conductivity of rock (ETCR) is widely used in the field of geology engineering. However, the ETCR is influenced by various factors, and the determination of ETCR is difficult. This paper presents a group method of data handling (GMDH) technique to predict the ETCR. Totally 270 measurements containing four input variables (i.e., temperature T, pressure P, porosity n and density ρ) were collected. A GMDH model of rock thermal conductivity was proposed based on this database, with R2 and RMSE values of 0.940 and 0.126. In addition, several existing models are employed to compare with the GMDH model. According to the results, these models have similar change tendency. However, these existing models can only reflect the effect of temperature on the ETCR, while the GMDH model can reflect the effects of more parameters. Moreover, a sensitivity analysis was performed. The results indicate that this model can well characterize the coupling relationship between the ETCR and two main parameters, i.e., temperature and pressure. The predicted ETCR increases with the increase of pressure, while it decreases with the increase of temperature. Finally, a graphical user interface (GUI) was established for practical application.

A simplified procedure for the prediction of liquefaction‐induced settlement of offshore wind turbines supported by suction caisson foundation based on effective stress analyses and an ML‐based group method of data handling

AbstractIn this research, a series of non‐linear dynamic finite element (FE) effective stress analyses were conducted to analyze the influence of the suction caisson geometry, ground motion intensity and contact pressure caused by the offshore wind turbine (OWT) on the settlement pattern and seismic demand of the OWT's structure on saturated dense sand. The baseline model and the FE procedure were validated using a database of well‐documented centrifuge tests. However, particular attention was given to the calibration campaign based on the measured system response quantities, such as the settlement, acceleration and pore‐pressure time histories. The FE results identified the contact pressure as an important state parameter caused by the OWT's mass; the governing ground‐shaking intensity measures that play a significant role in the derivation of an analytical framework for predicting liquefaction‐induced OWT settlements during major events are the shaking intensity rate (SIR), Arias Intensity ( and spectral acceleration at a period equal to 1 s (T = 1 s). The results revealed that approximating expressions derived using the modified least‐squares method (MLSM) reasonably capture the complex phenomenon of liquefaction‐induced settlement, with some exceptions at lower SIR values. Finally, to obtain the approximating expressions, the database was combined with a machine learning (ML)‐based group method of data handling (GMDH) that appropriately describes the interplay of multiple properties of the foundation soil, structure and seismic events while incorporating the effect of the interaction between the suction caisson, foundation soil, excess pore‐pressure generation and cyclic shear stresses.

Toward smart correlations for predicting in-situ stress: Application to evaluating subsurface energy structures

The precise calculation of the in-situ stress tensor is a crucial factor in addressing the challenges associated with the development of subsurface energy structures. To establish a consistent relationship between breakout shapes and the in-situ stress using nonlinear or coupled process assumptions through numerical methods, effective techniques are required. In this regard, three white box algorithms, namely gene expression programming (GEP), genetic programming (GP), and the group method of data handling (GMDH), were developed to predict the maximum horizontal stress. To develop a robust correlation using the white box algorithms, 662 data points were obtained from numerical analysis using an elastoplastic model across a wide range of wellbore pressures. Input parameters were breakout width and depth, wellbore pressure, and minimum horizontal stress. The study results indicated that the GP algorithm demonstrates higher accuracy compared to the GEP and GMDH, with a root mean square error (RMSE) of 0.9977 and a determination coefficient (R2) of 0.97564. Additionally, both SHAP values (SHapley Additive exPlanations) and sensitivity analysis were employed. The sensitivity analysis revealed that breakout width has a greater influence on predicting the maximum in-situ stress compared to other parameters. Furthermore, the Leverage technique indicated that the GP model can be considered a reliable tool for accurately estimating the in-situ stress, making it suitable for use in the subsurface energy structures.

Interfacial Tension-Temperature-Pressure-Salinity Relationship for the Hydrogen-Brine System under Reservoir Conditions: Integration of Molecular Dynamics and Machine Learning.

Hydrogen (H2) underground storage has attracted considerable attention as a potentially efficient strategy for the large-scale storage of H2. Nevertheless, successful execution and long-term storage and withdrawal of H2 necessitate a thorough understanding of the physical and chemical properties of H2 in contact with the resident fluids. As capillary forces control H2 migration and trapping in a subsurface environment, quantifying the interfacial tension (IFT) between H2 and the resident fluids in the subsurface is important. In this study, molecular dynamics (MD) simulation was employed to develop a data set for the IFT of H2-brine systems under a wide range of thermodynamic conditions (298-373 K temperatures and 1-30 MPa pressures) and NaCl salinities (0-5.02 mol·kg-1). For the first time to our knowledge, a comprehensive assessment was carried out to introduce the most accurate force field combination for H2-brine systems in predicting interfacial properties with an absolute relative deviation (ARD) of less than 3% compared with the experimental data. In addition, the effect of the cation type was investigated for brines containing NaCl, KCl, CaCl2, and MgCl2. Our results show that H2-brine IFT decreases with increasing temperature under any pressure condition, while higher NaCl salinity increases the IFT. A slight decrease in IFT occurs when the pressure increases. Under the impact of cation type, Ca2+ can increase IFT values more than others, i.e., up to 12% with respect to KCl. In the last step, the predicted IFT data set was used to provide a reliable correlation using machine learning (ML). Three white-box ML approaches of the group method of data handling (GMDH), gene expression programming (GEP), and genetic programming (GP) were applied. GP demonstrates the most accurate correlation with a coefficient of determination (R2) and absolute average relative deviation (AARD) of 0.9783 and 0.9767%, respectively.

Development of artificial intelligence-based models for prediction of vanadium adsorption onto activated carbon nanocomposites

Due to environmental concerns and economic value, the adsorption process is one of the suitable methods to separate vanadium from wastewater. In this study, two artificial intelligence models, including an adaptive neuro fuzzy inference system and group method of data handling, were successfully implemented to model the vanadium adsorption process with activated carbon nanocomposites. In the developed smart models, initial pH, contact time, and iron nanoparticle concentration were selected as input variables, and vanadium adsorption was chosen as the target parameter. In order to accomplish modeling, experimental data were extracted from our previous study and utilized to develop artificial intelligence-based models. For the neuro fuzzy approach, the average absolute relative deviation (AARD) and coefficient of determination (R2) were 2.45 % and 0.9858, respectively. The values of AARD and R2 for the group method of data handling technique were obtained 5.06 % and 0.9567, respectively. These results indicated the high accuracy of both intelligent models in predicting vanadium adsorption efficiency with activated carbon nanoparticles. Moreover, the comparison of the proposed smart approaches with the mechanism-based models, revealed the superiority of the neuro fuzzy model in predicting vanadium adsorption.

Modeling Hydrodynamic Behavior of the Ottawa River: Harnessing the Power of Numerical Simulation and Machine Learning for Enhanced Predictability

The Ottawa River Watershed is a vast area that stretches across Ontario and Quebec and holds great importance for Canada’s people, economy, and collective history, both in the present and the future. The river has faced numerous floods in recent years due to climate change. The most significant flood occurred in 2019, surpassing a 100-year flood event, and serves as a stark reminder of how climate change impacts our environment. Considering the limitations of machine learning (ML) models, which heavily rely on historical data used during training, they may struggle to accurately predict such “non-experienced” or “unseen” floods that were not encountered during the training process. To tackle this challenge, our study has utilized a combination of numerical modeling and ML to create an integrated methodology. Indeed, a comprehensive dataset of river flow discharge was generated using a numerical model, encompassing a wide range of potential future floods. This significantly improved the ML training process to generalize the accuracy of results. Utilizing this dataset, a novel ML model called the Expanded Framework of Group Method of Data Handling (EFGMDH) has been developed. Its purpose is to provide decision-makers with explicit equations for estimating three crucial hydrodynamic characteristics of the Ottawa River: floodplain width, flow velocity, and river flow depth. These predictions rely on various inputs, including the location of the desired cross-section, river slope, Manning roughness coefficient at different river sections (right, left, and middle), and river flow discharge. To establish practical models for each of the aforementioned hydrodynamic characteristics of the Ottawa River, different input combinations were tested to identify the most optimal ones. The EFGMDH model demonstrated high accuracy throughout the training and testing stages, achieving an R2 value exceeding 0.99. The proposed model’s exceptional performance demonstrates its reliability and practical applications for the study area.

Estimating discharge coefficient of triangular free overfall using the GMDH technique

Abstract Free overfalls are hydraulic structures used in flood control, water supply, irrigation, and flow measurements. The hydraulic systems of free overfall depend on rectangular end shape. The studies that dealt with triangular crest are few and almost non-existent. In this study, a triangular end-shape design uses multiple linear regression (MLR) and group method of data handling (GMDH) methods for four models with six sub-models. Then, 24 scenarios were chosen and compared. The discharge coefficient (Cd) of a free overfall with a triangular terminal was predicted using experimental data. The triangular end edge shape increased crest length, the discharge coefficient, and discharge passing over free overfall. To this goal, 180 triangular free fall tests were performed. Data were collected for two triangular free overfalls with an opposite flow direction with three angles 600, 750, and 900. Results of Cd acquired using the two ways discussed above show that the algorithm GMDH outperforms the other method. Values for the GMDH approach mod46 testing variables: RMSE, MARE, SI, R2, and NSE are 6.08 × 10−17, 2.65 × 10−17, 6.00 × 10−17, 100.00%, and 1.00, respectively, while these values for MLR are 0.06332, 0.05970, 0.06624, 15.431%, and −3.0419, respectively. The GMDH technique shows the best results concerning MLR and then chooses the best four scenarios from 24 with a Cd percentage error not exceeding ±2%.

Terrestrial water storage anomaly estimating using machine learning techniques and satellite‐based data (a case study of Lake Urmia Basin)

AbstractIn this study, the Terrestrial Water Storage Anomaly (TWSA) in the Lake Urmia Basin (LUB) was obtained by using the GRACE satellites. The whole study area was covered by 10 GRACE/GFO pixels, and the TWSA value was calculated for these 10 pixels. Examining the changing trend showed that the value of TWSA in the LUB has a decreasing trend and fluctuates from approximately −200 to +200 compared to the average value. The TWSA was modelled using the group method of data handling (GMDH) by considering six different parameters obtained from the European Centre for Medium‐Range Weather Forecasts (ECMWF) Reanalysis V5 (ERA5) and the Global Land Data Assimilation System version 2.0 (GLDAS V2.0). Finally, the best models with two to six input variables were selected. Among the models with different inputs, the GMDH Model 2, with three inputs, had the best performance compared to the other models. After choosing the best inputs in calculating the TWSA value by the GMDH, the TWSA value was also modelled by using the Adaptive Neuro‐Fuzzy Inference System (ANFIS), the Extreme Learning Machine (ELM) and the Artificial Neural Network (ANN). The results showed that GMDH not only outperformed other models but also provided some simple equations to apply in practical tasks.

Enhancing Flood Prediction Accuracy through Integration of Meteorological Parameters in River Flow Observations: A Case Study Ottawa River

Given that the primary cause of flooding in Ontario, Canada, is attributed to spring floods, it is crucial to incorporate temperature as an input variable in flood prediction models with machine learning algorithms. This inclusion enables a comprehensive understanding of the intricate dynamics involved, particularly the impact of heatwaves on snowmelt, allowing for more accurate flood prediction. This paper presents a novel machine learning approach called the Adaptive Structure of the Group Method of Data Handling (ASGMDH) for predicting daily river flow rates, incorporating measured discharge from the previous day as a historical record summarizing watershed characteristics, along with real-time data on air temperature and precipitation. To propose a comprehensive machine learning model, four different scenarios with various input combinations were examined. The simplest model with three parameters (maximum temperature, precipitation, historical daily river flow discharge) achieves high accuracy, with an R2 value of 0.985 during training and 0.992 during testing, demonstrating its reliability and potential for practical application. The developed ASGMDH model demonstrates high accuracy for the study area, with a significant number of samples having a relative error of less than 15%. The final ASGMDH-based model has only a second-order polynomial (AICc = 19,648.71), while it is seven for the classical GMDH-based model (AICc = 19,701.56). The sensitivity analysis reveals that maximum temperature significantly impacts the prediction of daily river flow discharge.