Figure 2 - available via license: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
Content may be subject to copyright.
| Relative errors computed for optimum rectangular channels predicted by ANN and GP: (a) y à and (b) bÃ.
Source publication
Lined channels with trapezoidal, rectangular and triangular sections are the most common manmade canals in practice. Since the construction cost plays a key role in water conveyance projects, it has been considered as the prominent factor in optimum channel designs. In this study, artificial neural networks (ANN) and genetic programming (GP) are us...
Citations
... The foundation of the artificial neural network (ANN) model is grounded in its nonlinear mapping structure, inspired by the configuration of human neurons. This model has proven effective in addressing diverse challenges across various fields of activity [7]. Within the domain of modeling methodologies, the adaptive neuro-fuzzy inference system (ANFIS) emerges as a sophisticated alternative, seamlessly blending the complexities of ANN with the nuanced architecture of a fuzzy inference system (FIS). ...
The availability of reliable electrical power, which is essential for a comfortable lifestyle worldwide, requires realistic power usage projections for electric utilities and policymakers, leading to the adoption of machine learning-based modelling tools due to the limitations of traditional power usage projection approaches. However, successful modeling of power usage in neuro-fuzzy models depends on the optimal selection of hyper-parameters. Consequently, this research looked at the major impact clustering methods and hyper-parameter modifications on a particle swarm optimization (PSO)-based adaptive neuro-fuzzy inference system (ANFIS) model. The study examined two distinct clustering methods and other key hyperparameters such as the number of clusters and cluster radius, resulting in a total of 10 sub-models. The performance of the developed models was assessed using four widely recognized performance indicators: root mean square error, mean absolute percentage error (MAPE), mean absolute error (MAE), and coefficient of variation of the root mean square error (CVRMSE). Additionally, the robustness of the optimal sub-model was evaluated by comparing it with other hybrid models based on three different PSO variants. The results revealed that the combination of the ANFIS approach and PSO, specifically with two clusters, yielded the most accurate forecasting scheme with the optimal values for MAPE (7.7778%), MAE (712.6094), CVRMSE (9.5464), and RMSE (909.4998).
... arameters. Unlike classical trend analysis methods, such as Mann-Kendall's test and Spearman's rho, it is not subject to constraints, like data length, independent structure of time series, and normality assumption. On the other hand, it is open to interpretation instead of the monotonous trend detection in classical methods (Şen 2012;Farrokhi (1) al. 2020;. Because of these advantages, it is a popular trend determination method and consequently used to quantify changes in meteorological parameters (Dabanlı et al. 2016;) and drought indexes (Caloiero 2018;Yilmaz 2019). ITA is basically based on marking the data in the Cartesian coordinate system (Fig. 2). For this purpose, the data length ...
... ANN is a widely used ML that resembles human brain (Niazkar 2020). In this study, the activation function for the hidden layer(s) was tanh, while the output layer was linear. ...
This study aims to assess the Eastern Black Sea Basin drought conditions. For this purpose, the trend changes in SPI values of 6, 9, 12, and 24 months using innovative trend analysis were examined. Additionally, four machine learning models, including Multiple Linear Regression, Artificial Neural Networks, K Nearest Neighbors, and XGBoost Regressor, are employed to forecast SPI with rainfall data between 1965 and 2020 from eight rainfall stations. The input data for each model was SPI values from lead times of 1 to 6, resulting into 768 unique scenarios. The ML models estimated SPI values better as the SPI duration increased, with the 24-month SPI showing the highest accuracy. The results of SPI forecast indicated that the optimal model and number of input variables varied for each SPI and station, indicating that further studies are required to improve SPI predictions.
... An artificial neural network (ANN) learns and transmits information from one artificial neuron to the next (Bisong 2019;Niazkar 2020). ANN models can handle problems that tolerate high error rates and have a large amount of example data (Nagy et al. 2002). ...
Estimating total sediment discharge is challenging. This study aims to assess performances of various data-driven models including empirical equations, machine learning (ML), and ensemble models for such estimations. The ML models include Support Vector Machine (SVM), Artificial Neural Network (ANN), K Nearest Neighbor (KNN), Random Forest (RF), and Decision Tree Regression (DTR). For this purpose, 543 widely-ranged data were collected from the United States Geological Survey (USGS) resources and used to train and test different models. Ranking different models demonstrated that Ackers and White's equation outperformed multiple linear regression (MLR) and SVM, which indicates that all ML models do not necessarily outperform empirical equations. Moreover, despite conducting multiple runs and parameter tuning, the results consistently indicated that increasing the number of hidden layers and neurons in ANN structures did not significantly improve the overall performance of the ANN models. In addition, the nonlinear ensemble model outperformed all methods and placed first in the ranking. Despite a notable difference between metrics obtained by KNN for the train and test data, it outperformed other methods and ranked second, while ANN achieved the third-best ranking place. The obtained result was also confirmed by the reliability analysis and confidence limits. However, due to negative predictions for some small sediment discharges by the nonlinear ensemble method, it did not demonstrate good reliability. Finally, the comparative analysis indicates that selecting a suitable model for estimating sediment discharges with a desirable accuracy is challenging, while further studies are required to assess other ML models or variants of ensemble models.
... For instance, multiple nonlinear regression modelling techniques, such as artificial neural networks (ANNs) and adaptive neuro-fuzzy inference systems (ANFISs), model experimental data by a combination of nonlinear parameters [6]. ANNs are a nonlinear mapping structure built from human neurons and have been used to address a wide range of problems in varied fields [7]. To increase modeling speed, fault tolerance, and addictiveness, ANFIS blends an ANN with a fuzzy inference system (FIS) [8]. ...
Electricity is undeniably one of the most crucial building blocks of high-quality life all over the world. Like many other African countries, Nigeria is still grappling with the challenge of the energy crisis. However, accurate prediction of electricity consumption is vital for the operation of electric utility companies and policymakers. In response, this study underlines the application of hybrid modelling techniques for the accurate prediction of electricity consumption, using Lagos districts, Nigeria, as a case study. To begin with, this research investigates the performance of three evolutionary algorithms — Particle Swarm Optimization (PSO), Genetic Algorithm (GA), and Differential Evolution (DE) — to optimize the parameters of adaptive network-based fuzzy inference systems (ANFIS). In addition, the impact of renowned clustering techniques such as grid partitioning (GP), fuzzy c-means (FCM), and subtractive clustering (SC) on other pivotal key hyperparameters of the ANFIS was examined and analyzed. Furthermore, the robustness of the optimal sub-model was evaluated by comparing it with other hybrid models that are based on six different variants of PSO. The efficacy of the proposed model was evaluated using four standard statistical measures. Finally, the results showed that the combination of the ANFIS approach and PSO under an SC approach and clustering radius of 0.6 delivered the best forecast scheme with the highest accuracy of the MAPE (8.8418%), the MAE (872.1784), the CVRMSE (10.7895), and the RMSE (1.0945E+03). The simulation results were analyzed and compared to other approaches, revealing that the suggested model is better.
... Although efforts have been made to improve the estimation of bridge afflux, there is still a need for more reliable and sustainable approaches developed by new and advanced powerful optimization algorithms, ML models, and other data mining techniques to achieve a higher accuracy [15]. Furthermore, despite the importance of bridge backwater in ensuring the safe design of piers and other hydraulic structures, few studies have focused on this issue in the literature, and the efficiency of ML models in addressing this issue has not been adequately assessed. ...
Bridges are essential structures that connect riverbanks and facilitate transportation. However, bridge piers and abutments can disrupt the natural flow of rivers, causing a rise in water levels upstream of the bridge. The rise in water levels, known as bridge backwater or afflux, can threaten the stability or service of bridges and riverbanks. It is postulated that applications of estimation models with more precise afflux predictions can enhance the safety of bridges in flood-prone areas. In this study, eight machine learning (ML) models were developed to estimate bridge afflux utilizing 202 laboratory and 66 field data. The ML models consist of Support Vector Regression (SVR), Decision Tree Regressor (DTR), Random Forest Regressor (RFR), AdaBoost Regressor (ABR), Gradient Boost Regressor (GBR), eXtreme Gradient Boosting (XGBoost) for Regression (XGBR), Gaussian Process Regression (GPR), and K-Nearest Neighbors (KNN). To the best of the authors’ knowledge, this is the first time that these ML models have been applied to estimate bridge afflux. The performance of ML-based models was compared with those of artificial neural networks (ANN), genetic programming (GP), and explicit equations adopted from previous studies. The results show that most of the ML models utilized in this study can significantly enhance the accuracy of bridge afflux estimations. Nevertheless, a few ML models, like SVR and ABR, did not show a good overall performance, suggesting that the right choice of an ML model is important.
... Various statistical indices have been used to evaluate models, and each has a different relationship to express the error of the observed and predicted values [48]. Error evaluation measures are calculated for the training and test data. ...
Excessive population growth and high water demands have significantly increased water extractions from deep and semi-deep wells in the arid regions of Iran. This has negatively affected water quality in different areas. The Water Quality Index (WQI) is a suitable tool to assess such impacts. This study used WQI and the fuzzy hierarchical analysis process of the water quality index (FAHP-WQI) to investigate the water quality status of 96 deep agricultural wells in the Yazd-Ardakan Plain, Iran. Calculating the WQI is time-consuming, but estimating WQI is inevitable for water resources management. For this purpose, three Machine Learning (ML) algorithms, namely, Gene Expression Programming (GEP), M5P Model tree, and Multivariate Adaptive Regression Splines (MARS), were employed to predict WQI. Using Wilcox and Schoeller charts, water quality was also investigated for agricultural and drinking purposes. The results demonstrated that 75% and 33% of the study area have good quality, based on the WQI and FAHP-WQI methods, respectively. According to the results of the Wilcox chart, around 37.25% of the wells are in the C3S2 and C3S1 classes, which indicate poor water quality. Schoeller’s diagram placed the drinking water quality of the Yazd-Ardakan plain in acceptable, inadequate, and inappropriate categories. Afterwards, WQI, predicted by means of ML models, were compared on several statistical criteria. Finally, the comparative analysis revealed that MARS is slightly more accurate than the M5P model for estimating WQI.
... In addition, six metrics were used on the evaluating dataset to evaluate the performance of XGBoost and other state-of-the-art predictive models, including correlation coefficient (R), root mean square error (RMSE), mean absolute error (MAE), mean absolute relative error (MARE), Nash-Sutcliffe efficiency coefficient (NSE), and accuracy (ACC). These indicators are calculated as follows [41]: ...
As one of the physical quantities concerned in agricultural production, soil moisture can effectively guide field irrigation and evaluate the distribution of water resources for crop growth in various regions. However, the spatial variability of soil moisture is dramatic, and its time series data are highly noisy, nonlinear, and nonstationary, and thus hard to predict accurately. In this study, taking Jiangsu Province in China as an example, the data of 70 meteorological and soil moisture automatic observation stations from 2014 to 2022 were used to establish prediction models of 0–10 cm soil relative humidity (RHs10cm) via the extreme gradient boosting (XGBoost) algorithm. Before constructing the model, according to the measured soil physical characteristics, the soil moisture observation data were divided into three categories: sandy soil, loam soil, and clay soil. Based on the impacts of various factors on the soil water budget balance, 14 predictors were chosen for constructing the model, among which atmospheric and soil factors accounted for 10 and 4, respectively. Considering the differences in soil physical characteristics and the lagged effects of environmental impacts, the best influence times of the predictors for different soil types were determined through correlation analysis to improve the rationality of the model construction. To better evaluate the importance of soil factors, two sets of models (Model_soil&atmo and Model_atmo) were designed by taking soil factors as optional predictors put into the XGBoost model. Meanwhile, the contributions of predictors to the prediction results were analyzed with Shapley additive explanation (SHAP). Six prediction effect indicators, as well as a typical drought process that happened in 2022, were analyzed to evaluate the prediction accuracy. The results show that the time with the highest correlations between environmental predictors and RHs10cm varied but was similar between soil types. Among these predictors, the contribution rates of maximum air temperature (Tamax), cumulative precipitation (Psum), and air relative humidity (RHa) in atmospheric factors, which functioned as a critical factor affecting the variation in soil moisture, are relatively high in both models. In addition, adding soil factors could improve the accuracy of soil moisture prediction. To a certain extent, the XGBoost model performed better when compared with artificial neural networks (ANNs), random forests (RFs), and support vector machines (SVMs). The values of the correlation coefficient (R), root mean square error (RMSE), mean absolute error (MAE), mean absolute relative error (MARE), Nash–Sutcliffe efficiency coefficient (NSE), and accuracy (ACC) of Model_soil&atmo were 0.69, 11.11, 4.87, 0.12, 0.50, and 88%, respectively. This study verified that the XGBoost model is applicable to the prediction of soil moisture at the provincial level, as it could reasonably predict the development processes of the typical drought event.
... The MHBMO algorithm has been previously applied to solve various problems in water resources management (Niazkar, 2020). In essence, it is a zero-order search-based optimization algorithm that basically mimics the nature mating process of honey bees, whose community consists of the queen, drones, and workers, while the first two are the best solution in each generation and possible random values for calibration coefficients, i.e., a and b in Eq. (1) and c, d, e, and f in Eq. (2). ...
Sediment ratings supply an important input to the design of water resources projects. Nevertheless, the accuracy of sediment ratings has remained a matter of concern for hydrologists. The present article investigates both the aspect of improving the accuracy, i.e., modifying the simple rating curve equation by introducing a four-parameter equation and application of ensemble machine learning (ML) and ensemble empirical models, to estimate sediment loads. The ML models include artificial neural networks, multi-gene genetic programming (MGGP), and a hybrid MGGP-based model. Published field data at two measuring stations were used to assess the performance of different models employed in this study. The comparative analysis conducted in this study provides a novel comparison of sediment load estimations for three time scales. For instance, the ML-based simple average ensemble model (i.e., 556.5, 255.0, and 0.759) and the empirical-based nonlinear ensemble model (i.e., 549.1, 378.6, and 0.589) achieved the lowest root-mean-square errors and mean absolute errors and highest determination coefficients for the train and test monthly sediment data of the first station, respectively. Finally, the findings demonstrate that ensemble-based models generally improve the estimates of sediment loads at daily, 10-daily, and monthly scales.
... An example of a nonlinear forecasting model is the artifcial neural network (ANN). Te ANN model is a nonlinear mapping structure built on the framework of human neurons and has been efcaciously used to address a range of problems in various felds [24]. ANFIS is another modelling technique which amalgamates the ANN with a fuzzy inference system (FIS) to improve the speed, fault tolerance, and addictiveness of the modelling system [25]. ...
Future energy planning relies on understanding how much energy is produced and consumed. In response, this study developed a multihybrid adaptive neuro-fuzzy inference system (ANFIS) for students’ residences, using the University of Johannesburg residence, South Africa as a case study. The model input variables are wind speed, temperature, and humidity, with the output being the equivalent energy consumption for the student housing. While the particle swarm optimization (PSO) technique is versatile and widely used, it falls short by exhibiting premature convergence. To address this problem, the velocity update equation of the original PSO algorithm is modified by incorporating a dynamic linear decreasing inertia weight, which improves the PSO algorithm’s convergence behaviour and aids both local and global search. Following that, the modified PSO (MPSO) is used to optimize the ANFIS parameters for the best model prediction. A comparative analysis is conducted between the MPSO, the original PSO, and six other hybrid models using a dataset division of 70% for training and 30% for testing. Performance evaluation was carried out using three well-known performance benchmarks: root mean square error (RMSE), mean absolute deviation (MAD), and coefficient of variation (RCoV). The experimental results show that the performance of the proposed MPSO-ANFIS outperformed other methods with the least values of the RMSE (1.8928 KWh), MAD (1.5051 KWh), and RCoV (0.1370), respectively. Furthermore, when compared to the PSO-ANFIS, the MPSO-ANFIS demonstrated improvements in RMSE, MAD, and RCoV with 1.58%, 2.11%, and 5.23%, respectively. Based on the results, it can be concluded that the MPSO-ANFIS provides better prediction accuracy which is vital for strategic energy planning.
... Generally, GP is an AI technique with a tree-like flexible structure (Niazkar and Zakwan, 2021a). It basically exploits GA as a search engine to seek for a suitable relationship that converts a set of input data into output data (Niazkar, 2020). In other words, GP tackled the mentioned problem by solving an optimization problem using GA, while the objective function is to either minimize or maximize the error between the estimated and real output data. ...
... Based on the current literature, MGGP can be used as a prediction tool for many problems, which the AI models except MGGP has been used. Several examples of these problems include developing bed roughness predictors (Giustolisi, 2004;Niazkar et al., 2019a), design of open channels (Niazkar, 2020), predicting scour depth around piers (Niazkar and Afzali, 2018), estimating water surface profiles , and hydrological flood and stage routing (Sivapragasam et al., 2008;Fallah-Mehdipour et al., 2013). ...
