Search

Abstract Groundwater level fluctuations are one of the main components of the hydrogeological cycle and one of the required variables for many water resources operation models. The numerical models can estimate groundwater level (GWL) based on extensive statistics and information and using complex equations in any area. But one of the most important challenges in analyzing and predicting groundwater depletion in water management is the lack of reliable and complete data. For this reason, the use of artificial intelligence models with high predictive accuracy and due to the need for less data is inevitable. In recent years, the use of different numerical models has been noticed as an efficient solution. These models are able to estimate groundwater levels in any region based on extensive statistics and information and also various field experiments such as pumping tests, geophysics, soil and land use maps, topography and slope data, different boundary conditions and complex equations. In the current research, first, by using available statistics, information and maps, the groundwater level fluctuations of the Sonqor plain are simulated by the GMS model, and the accuracy of the model is evaluated in two stages of calibration and validation. Then, due to the need for much less data volume in artificial intelligence-based methods, the GA-ANN and ICA-ANN hybrid methods and the ELM and ORELM models are utilized. The results display that the output of the ORELM model has the best fit with observed data with a correlation coefficient equal to 0.96, and it also has the best and closest scatter points around the 45 degrees line, and in this sense, it is considered as the most accurate model. To ensure the correct selection of the best model, the Taylor diagram is also used. The results demonstrate that the closest point to the reference point is related to the ORELM method. Therefore, to predict the groundwater level in the whole plain, instead of using the complex GMS model with a very large volume of data and also the very time-consuming process of calibration and verification, the ORELM model can be used with confidence. This approach greatly helps researchers to predict groundwater level variations in dry and wet years using artificial intelligence with high accuracy instead of numerical models with complex and time-consuming structures.

River streamflow is an essential hydrological parameters for optimal water resource management. This study investigates models used to estimate monthly time-series river streamflow data at two hydrological stations in the USA (Heise and Irwin on Snake River, Idaho). Five diverse types of machine learning (ML) model were tested, support vector machine-radial basis function (SVM-RBF), SVM-Polynomial (SVM-Poly), decision tree (DT), gradient boosting (GB), random forest (RF), and long short-term memory (LSTM). These were trained and tested alongside a conventional multiple linear regression (MLR). To improve the estimation and model performance, hybrid models were designed by coupling the models with wavelet theory (W). The models performance was assessed using root mean square error (RMSE), mean absolute error (MAE), coefficient of determination (R2), Nash-Sutcliffe efficiency (NSE), and Willmott’s index (WI). A side-by-side performance assessment of the stand-alone and hybrid models revealed that the coupled models exhibit better estimates of monthly river streamflow relative to the stand-alone ones. The statistical parameter values for the best model (W-LSTM4) during the test phase was RMSE = 36.533 m3/s, MAE = 26.912 m3/s, R2 = 0.947, NSE = 0.946, WI = 0.986 (Heise station), and RMSE = 33.378 m3/s, MAE = 24.562 m3/s, R2 = 0.952, NSE = 0.951, WI = 0.987 (Irwin station).

The techniques employed to manage an industrial project are based on tools that aim to achieve the objectives set by an organization. Most of these techniques consider the development of operative and predictive models. The difficulty in developing project planning models relies on estimating large sets of parameters and the need to include model sections of poorly identifiable, that increase costs and time. This work develops a hybrid forecasting model for all the phases that make up die-casting projects through a series of parameters and sub-models that contemplate the particularities of each case, thereby achieving greater precision in the forecast. The model identifies the cost and time factors that affect project planning, specifically in the die-casting industry, and intends to predict their future behaviour when certain initially given conditions are modified. To estimate the parameters of the hybrid model, several factors in the processes were considered that interact in this industry, such as primary matter costs and activities associated to the process. The considered processes that have a substantial economic impact on the implementation of the project were selected. The criteria for this selection considered identifying the relevant parts of the design and manufacturing in the diecasting industry. Process factors such as the Cost of aluminium and its related activities, whose processes will be grouped into cost and time entities to build a set of metrics that allow better control over them. Finally, the proposed model is based on analytical, parametric, and analog methods that achieve accuracy greater than 85 % in predicting the time and Cost of the process. [ABSTRACT FROM AUTHOR]

A reliable prediction of the spatio-temporal drought variation can lead to a reduction in vulnerability and an improvement in the management of drought-dependent businesses. In this study, three clustering algorithms, K-mean, Hierarchical and Expectation–Maximization, were first used to divide Southern Italy into homogeneous drought regions, based on gridded data of the Standardized Precipitation Evapotranspiration Index forecasting with a 6 months' time scale (SPEI6). The Hierarchical algorithm identified five well-distinct clusters characterized by drought events of different duration and severity, considering the different morphoclimatic characteristics of the study area. Then, the mean SPEI6 time series were evaluated for each cluster and used to assess the evolutionary drought trends. In addition, two Machine Learning (ML) algorithms, M5P and Support Vector Regression (SVR), were also used to develop forecasting models for the SPEI6, without the need for additional exogenous inputs. Moreover, the Stacking ML technique was used to develop a hybrid model based on both individual M5P and SVR algorithms. The clustering-forecasting combination makes it possible to identify the evolutionary trends taking place in the various homogeneous areas in a concise but effective manner. The hybrid M5P-SVR model (R2 up to 0.91, minimum Root Mean Square Error RMSE = 0.38) outperformed both M5P (R2 up to 0.87, minimum RMSE = 0.42) and SVR (R2 up to 0.89, minimum RMSE = 0.39) models, showing to be particularly suitable for drought forecasting in areas with long and severe drought events. [ABSTRACT FROM AUTHOR]

Droughts cause severe damage to the economy, society, and environment. Drought forecasting plays an important role in establishing mitigation drought damage plans. In this study, a hybrid model involving long short-term memory and a climate model (LSTM-CM) is constructed for drought prediction. LSTM-CM was compared to the long short-term model stand-alone (LSTM-SA) and climate prediction model GloSea5 (GS5). The performance of models was evaluated based on the Pearson correlation coefficient (CC), mean absolute error (MAE), root mean squared error (RMSE), and skill score (SS). GS5 displayed physical robustness in predictions and did not reduce the amplitude or shift results. However, GS5 prediction tends to have a large bias caused by the inputs, model structure, and parameters. The MAEs of GS5 at 1, 2 and 3 months (0.41, 0.68, and 0.89) were higher than those of LSTM-SA (0.38, 0.61, and 0.89). The LSTM-SA reduced bias, but predictions were characterized by shifts, small variance, and failure to capture drought occurrences in long-lead-time cases. LSTM-CM yielded enhanced drought predictions by encompassing the low bias of LSTM-SA and the physical process simulation ability of GS5; thus, it inherited the good features of these models and limited the poor features. The SS values based on the CC, MAE, and RMSE of LSTM-CM compared to those of GS5 for 1-, 2-, and 3-month lead time predictions were improved from 29.17 to 54.29, 22.47 to 34.15, and 1.75 to 35.09%, respectively. LSTM-CM can accurately detect drought events and displayed less uncertainty in prediction than LSTM-SA and GS5. [ABSTRACT FROM AUTHOR]

During the recent few decades, the use of various models has been regarded as a promising option to predict groundwater level (GWL) in any given region using a wide variety of data and relevant equations. The lack of trustworthy and comprehensive data is, nevertheless, one of the most significant obstacles that must be overcome in order to analyze and anticipate the depletion of groundwater in the context of water management. Because of this, the implementation of artificial intelligence (AI) models that are able to predict the GWL with high accuracy using a reduced amount of data is unavoidable. In this work, the GWL variations of Lur plain were simulated using GMS model by utilizing the available data and maps. The accuracy of model was assessed at both phases i.e. validation and calibration. Following that, GA-ANN and ICA-ANN approaches, together with ELM, ORELM, and GMDH models, were used in order to fulfill the demand for too smaller volumes by AI procedures. According to the results, the ORELM output had the highest correlation with the observed information, which indicates that it is the most accurate model in this regard. The correlation coefficient for this model was 0.976. Because of this, instead of utilizing a complicated GMS model that needs a significant amount of data for the simulation, an ORELM model can be used to reliably forecast the GWL in the Lur plain. This simple model allows the researchers to accurately predict changes in GWL during rainy and non-rainy years compared to other complicated and time-consuming numerical models. [ABSTRACT FROM AUTHOR]

This study analyzes the transmission of market uncertainty on key European financial markets and the cryptocurrency market over an extended period, encompassing the pre-, during, and post-pandemic periods. Daily financial market indices and price observations are used to assess the forecasting models. We compare statistical, machine learning, and deep learning forecasting models to evaluate the financial markets, such as the ARIMA, hybrid ETS-ANN, and kNN predictive models. The study results indicate that predicting financial market fluctuations is challenging, and the accuracy levels are generally low in several instances. ARIMA and hybrid ETS-ANN models perform better over extended periods compared to the kNN model, with ARIMA being the best-performing model in 2018–2021 and the hybrid ETS-ANN model being the best-performing model in most of the other subperiods. Still, the kNN model outperforms the others in several periods, depending on the observed accuracy measure. Researchers have advocated using parametric and non-parametric modeling combinations to generate better results. In this study, the results suggest that the hybrid ETS-ANN model is the best-performing model despite its moderate level of accuracy. Thus, the hybrid ETS-ANN model is a promising financial time series forecasting approach. The findings offer financial analysts an additional source that can provide valuable insights for investment decisions. [ABSTRACT FROM AUTHOR]

Recognition of handwritten Arabic characters remains a major challenge for researchers, given the significant differences in handwriting. This paper presents a hybrid method based on combining the most efficient classification techniques. A trained convolutional neural network (CNN) was applied to extract features from character images. Then, a support vector machine (SVM) was used for classification. By combining CNN and SVM, the aim is to exploit both technologies' strengths. Four hybrid models are proposed in this work. Several databases such as HACDB, HIJJA, AHCD, and MNIST were used to evaluate them. The results obtained are satisfactory compared to similar studies in the literature, with a test accuracy of 89.7, 88.8, 97.3, and 99.4%, respectively. [ABSTRACT FROM AUTHOR]

Short-term load forecasting (STLF) is critical for the energy industry. Accurate predictions of future electricity demand are necessary to ensure power systems' reliable and efficient operation. Various STLF models have been proposed in recent years, each with strengths and weaknesses. This paper comprehensively reviews some STLF models, including time series, artificial neural networks (ANNs), regression-based, and hybrid models. It first introduces the fundamental concepts and challenges of STLF, then discusses each model class's main features and assumptions. The paper compares the models in terms of their accuracy, robustness, computational efficiency, scalability, and adaptability and identifies each approach's advantages and limitations. Although this study suggests that ANNs and hybrid models may be the most promising ways to achieve accurate and reliable STLF, additional research is required to handle multiple input features, manage massive data sets, and adjust to shifting energy conditions. [ABSTRACT FROM AUTHOR]

With the constantly growing number of reviews and other sentiment-bearing texts on the Web, the demand for automatic sentiment analysis algorithms continues to expand. Aspect-based sentiment classification (ABSC) allows for the automatic extraction of highly fine-grained sentiment information from text documents or sentences. In this survey, the rapidly evolving state of the research on ABSC is reviewed. A novel taxonomy is proposed that categorizes the ABSC models into three major categories: knowledge-based, machine learning, and hybrid models. This taxonomy is accompanied with summarizing overviews of the reported model performances, and both technical and intuitive explanations of the various ABSC models. State-of-th-eart ABSC models are discussed, such as models based on the transformer model, and hybrid deep learning models that incorporate knowledge bases. Additionally, various techniques for representing the model inputs and evaluating the model outputs are reviewed. Furthermore, trends in the research on ABSC are identified and a discussion is provided on the ways in which the field of ABSC can be advanced in the future. [ABSTRACT FROM AUTHOR]

Sentiment classification plays a pivotal role in natural language processing (NLP), and prior research has established the efficacy of utilizing convolutional neural networks (CNNs) and long short-term memory (LSTM) in this task. However, these approaches suffer from individual performance limitations: CNNs are limited to extracting local information and fail to express context information adequately, while LSTM networks excel at extracting context dependencies but exhibit long training times. To address this issue, we propose a novel text classification algorithm based on a hybrid CNN-LSTM model that leverages the strengths of both approaches and overcomes their limitations by combining them. Our approach is evaluated on the IMDB dataset, and we present a hyperparameter optimization framework utilizing Random Search to increase the likelihood of producing an optimally performing model. [ABSTRACT FROM AUTHOR]

Accurate prediction of stock trends is of great crucial to guiding investment management and financial policymaking. In recent years, text content generated by investors on various media platforms has had a big impact on stock movements. However, most of the existing studies apply equal weight to the text content to construct the sentiment index and ignore the fact that the influence of the sentiment index on the stock market will decrease with the increase of time interval. In this paper, we propose a model based on dual classifier coupling and sentiment analysis to study the trend of stock index. (1) We come up with a sentiment index weighted based on the reading volume to predict the trend of the stock index. (2) The exponentially weighted moving average (EWMA) model is used to modify the weighted sentiment index to obtain the modified sentiment index. (3) The maximum information coefficient (MIC) is applied to calculate the correlation between the closing price of the stock index and the modified sentiment index to obtain the best-modified sentiment index. (4) Two frequently used classifiers, convolutional neural network (CNN) and support vector machine (SVM), are used to build a dual-classifier coupling prediction model and adopt it for the final classification prediction. Two Chinese stock indexes (SSE 50 and CSI 300) are used to evaluate the predictive effect of the proposed model. In the practice case of the SSE 50 Index, the precision, accuracy, recall rate, F1-score, and AUC reached 81.38%, 80.99%, 80.99%, 81.04%, and 81.22%, respectively, after adding the modified sentiment index. In the practical case of the CSI 300 Index, the precision, accuracy, recall rate, F1-score, and AUC reached 80.07%, 80.07%, 80.07%, 80.06%, and 80.03%, respectively, after adding the modified sentiment index. It can be found through the experimental results that adding the investor sentiment index can advance the trend prediction of the stock index, and adding the modified sentiment index has a more obvious improvement effect. After adding the modified investor sentiment index, the prediction results of our proposed two-classifier coupled CNN-SVM model are much better than other benchmark models. [ABSTRACT FROM AUTHOR]

Predicting reservoir water levels helps manage droughts and floods. Predicting reservoir water level is complex because it depends on factors such as climate parameters and human intervention. Therefore, predicting water level needs robust models. Our study introduces a new model for predicting reservoir water levels. An extreme learning machine, the multi-kernel least square support vector machine model (MKLSSVM), is developed to predict the water level of a reservoir in Malaysia. The study also introduces a novel optimization algorithm for selecting inputs. While the LSSVM model may not capture nonlinear components of the time series data, the extreme learning machine (ELM) model—MKLSSVM model can capture nonlinear and linear components of the time series data. A coati optimization algorithm is introduced to select input scenarios. The MKLSSVM model takes advantage of multiple kernel functions. The extreme learning machine model—multi-kernel least square support vector machine model also takes the benefit of both the ELM model and MKLSSVM model models to predict water levels. This paper's novelty includes introducing a new method for selecting inputs and developing a new model for predicting water levels. For water level prediction, lagged rainfall and water level are used. In this study, we used extreme learning machine (ELM)-multi-kernel least square support vector machine (ELM-MKLSSVM), extreme learning machine (ELM)-LSSVM-polynomial kernel function (PKF) (ELM-LSSVM-PKF), ELM-LSSVM-radial basis kernel function (RBF) (ELM-LSSVM-RBF), ELM-LSSVM-Linear Kernel function (LKF), ELM, and MKLSSVM models to predict water level. The testing means absolute of the same models was 0.710, 0.742, 0.832, 0.871, 0.912, and 0.919, respectively. The Nash–Sutcliff efficiency (NSE) testing of the same models was 0.97, 0.94, 0.90, 0.87, 0.83, and 0.18, respectively. The ELM-MKLSSVM model is a robust tool for predicting reservoir water levels. [ABSTRACT FROM AUTHOR]

The recognition of emotions is one of the most challenging issues in human–computer interaction (HCI). EEG signals are widely adopted as a method for recognizing emotions because of their ease of acquisition, mobility, and convenience. Deep neural networks (DNN) have provided excellent results in emotion recognition studies. Most studies, however, use other methods to extract handcrafted features, such as Pearson correlation coefficient (PCC), Principal Component Analysis, Higuchi Fractal Dimension (HFD), etc., even though DNN is capable of generating meaningful features. Furthermore, most earlier studies largely ignored spatial information between the different channels, focusing mainly on time domain and frequency domain representations. This study utilizes a pre-trained 3D-CNN MobileNet model with transfer learning on the spatio-temporal representation of EEG signals to extract features for emotion recognition. In addition to fully connected layers, hybrid models were explored using other decision layers such as multilayer perceptron (MLP), k-nearest neighbor (KNN), extreme learning machine (ELM), XGBoost (XGB), random forest (RF), and support vector machine (SVM). Additionally, this study investigates the effects of post-processing or filtering output labels. Extensive experiments were conducted on the SJTU Emotion EEG Dataset (SEED) (three classes) and SEED-IV (four classes) datasets, and the results obtained were comparable to the state-of-the-art. Based on the conventional 3D-CNN with ELM classifier, SEED and SEED-IV datasets showed a maximum accuracy of 89.18% and 81.60%, respectively. Post-filtering improved the emotional classification performance in the hybrid 3D-CNN with ELM model for SEED and SEED-IV datasets to 90.85% and 83.71%, respectively. Accordingly, spatial-temporal features extracted from the EEG, along with ensemble classifiers, were found to be the most effective in recognizing emotions compared to state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Photovoltaic (PV) power shows different fluctuation characteristics under different weather types as well as strong randomness and uncertainty in changeable weather such as sunny to cloudy, cloudy to rain, and so on, resulting in low forecasting accuracy. For the changeable type of weather, an ultra-short-term photovoltaic power forecasting method is proposed based on affinity propagation (AP) clustering, complete ensemble empirical mode decomposition with an adaptive noise algorithm (CEEMDAN), and bi-directional long and short-term memory network (BiLSTM). First, the PV power output curve of the standard clear-sky day was extracted monthly from the historical data, and the photovoltaic power was normalized according to it. Second, the changeable days were extracted from various weather types based on the AP clustering algorithm and the Euclidean distance by considering the mean and variance of the clear-sky power coefficient (CSPC). Third, the CEEMDAN algorithm was further used to decompose the data of changeable days to reduce its overall non-stationarity, and each component was forecasted based on the BiLSTM network, so as to obtain the PV forecasting value in changeable weather. Using the PV dataset obtained from Alice Springs, Australia, the presented method was verified by comparative experiments with the BP, BiLSTM, and CEEMDAN-BiLSTM models, and the MAPE of the proposed method was 2.771%, which was better than the other methods. [ABSTRACT FROM AUTHOR]

Landslide susceptibility mapping (LSM) is essential for land-use planning, as it helps to identify areas at risk of landslides and enables effective prevention measures to be taken. Various statistical and machine learning (ML) models are used in LSM, including SVM and ANFIS, which have shown promising results. However, determining which model performs better remains a key challenge. To address this issue, this paper aims to compare six hybrid models constructed with two well-known and powerful ML models, namely SVM and ANFIS, and three meta-heuristic algorithms, namely Genetic Algorithm (GA), Differential Evolution (DE), and Cultural Algorithm (CA), for LSM in a case study in western Serbia. In the process of building the models, 359 landslide sites and 14 determinants were used. The accuracy of the models was evaluated using several indexes, including Root Mean-Squared Error (RMSE), coefficient of determination (R2), and Area under the Receiver Operating Characteristic Curve (AUROC). The modeling results showed that the SVM-GA model has the highest accuracy (AUROC = 0.78) in predicting landslide incidence, followed by the ANFIS-GA (AUROC = 0.775), SVM-CA (AUROC = 0.773), ANFIS-DE (AUROC = 0.771), SVM-DE (AUROC = 0.76), and ANFIS-CA (AUROC = 0.65) models in validation phase. Therefore, the study suggested that SVM-based hybrid models are more accurate than ANFIS-based models for LSM, and thus, modelers may use SVM-based hybrid models for such applications. This study provides valuable insights into identifying the most appropriate and effective models for LSM, which can help to mitigate the risks associated with landslides and ensure sustainable land-use practices.

Identifying flood-prone regions is critical for effective management of flood hazards as floods are among the most devastating natural disasters globally. However, accurate modeling and prediction of floods are challenging due to their complexity. The current research has proposed a novel approach for Flood Hazard (FH) prediction using hybrid Machine Learning (ML) models that integrate ensemble ML models with several Feature Selection (FS) algorithms. An optimum set of Flood Influential Factors (FIFs) was determined using the Simulated Annealing (SA) and Information Gain (IG) FS algorithms. The ensemble ML models employed include AdaboostM1 (ABM), Boosted Generalized Linear Model (BGLM), and Stochastic Gradient Boosting (SGB) algorithms. In addition, the hyper-parameters of the hybrid models were optimized using the random search (RS) method and repeated cross-validation technique. The proposed hybrid models were trained using flood inventory map and FIFs obtained from a spatial database. The results verified that the SA and IG algorithms detect 9 and 13 factors as FIFs in the FH assessment, respectively. Moreover, rainfall, distance to river, altitude, and lithology FIFs have a greater impact than the other factors in the Sardabroud watershed, Mazandaran Province, Iran. Several robust indicators, such as the area under curve (AUC) in relative operating characteristic (ROC) curves and statistical measurements, were employed to assess the robustness of hybrid models. SA-ABM model had the highest AUC value (0.983), while IG-ABM, SA-BGLM, SA-SGB, IG-BGLM, and IG-SGB had lower values (0.952 to 0.973). Finally, the SA-ABM hybrid model classified 27% of the study area as having a high hazard of floods.

Flash floods are among the world most destructive natural disasters, and developing optimum hybrid Machine Learning (ML) models for flash flood susceptibility (FFS) modeling remains a challenge. This study proposed novel intelligence algorithms based on a hybrid of several ensemble ML models (i.e., Bagged Flexible Discriminant Analysis (BAFDA), Extreme Gradient Boosting (XBG), Rotation Forest (ROF) and Boosted Generalized Additive Model (BGAM)) and wrapper-based factor optimization algorithms (i.e., Recursive Feature Elimination (RFE) and Boruta) to improve the accuracy of FFS mapping at Neka-Haraz watershed in Iran. In addition, the Random Search (RS) method is proposed for meta-optimization of the developed hybrid models hyper-parameters. This study considers 20 flash flood conditioning factors (CgFs) and 380 flood and non-flood locations to create a geospatial database. The performance of each hybrid model was evaluated by area under the receiver operating characteristic (ROC) curve (AUC) and several validation methods, such as efficiency. The developed hybrid models demonstrated good performance, with BGAM-Boruta achieving the highest performance (AUC = 0.953, and Efficiency = 0.910), followed by ROF-Boruta (AUC = 0.952), ROF-RFE (AUC = 0.951), BAFDA-Boruta (AUC = 0.950), BGAM-RFE (AUC = 0.950), ROF (AUC = 0.949), BGAM (AUC = 0.948), BAFDA-RFE (AUC = 0.943), XGB-Boruta (AUC = 0.943), BAFDA (AUC = 0.939), XGB-RFE (AUC = 0.938) and XGB (AUC = 0.911). In the BGAM-Boruta model, the regional coverage was about 46% for high to very high FFS areas. Moreover, the study revealed that distance to river, slope, rainfall, altitude, and distance to road CgFs are the most significant for FFS modeling in this region.

The greenhouse effect formed by the massive emission of carbon dioxide has caused serious harm to the Earth’s environment, in which the power sector constitutes one of the primary contributors to global greenhouse gas emissions. Reducing carbon emissions from electricity plays a pivotal role in minimizing greenhouse gas emissions and mitigating the ecological, economic, and social impacts of climate change, while carbon emission prediction provides a valuable point of reference for the formulation of policies to reduce carbon emissions from electricity. The article provides a detailed review of research results on deep learning-based carbon emission prediction. Firstly, the main neural networks applied in the domain of carbon emission forecasting at home and abroad, as well as the models combining other methods and neural networks, are introduced, and the main roles of different methods, when combined with neural networks, are discussed. Secondly, neural networks were used to predict electricity carbon emissions, and the performance of different models on carbon emissions was compared. Finally, the application of neural networks in the realm of the prediction of carbon emissions is summarized, and future research directions are discussed. The article provides a reference for researchers to understand the research dynamics and development trend of deep learning in the realm of electricity carbon emission forecasting.

Accurate water level prediction is of great importance for water infrastructures such as dams, embankments, and agriculture. However, the water level has nonlinear characteristics, which make it very challenging to accurately predict the water level. This study proposes a combined model based on variational mode decomposition (VMD), a genetic algorithm–the ELMAN neural network–VMD–the autoregressive integrated moving average (ARIMA) model (GA–ELMAN–VMD–ARIMA). Firstly, VMD preprocesses the original water level and predicts each subsequence with the GA–ELMAN model. Then the error sequence is decomposed by VMD and predicted by the ARIMA model. Finally, the predicted water level is corrected. Using three groups of data from different sites, 10 models are established to compare the performance of the model. The results show that the combination of the VMD algorithm and the GA–ELMAN model can improve the performance of prediction on datasets. In addition, it also shows that the VMD double processing can greatly improve the prediction accuracy. HIGHLIGHTS The variational mode decomposition (VMD) double processing is used for water level prediction.; A genetic algorithm is used to optimize the hyperparameters of the ELMAN neural network.; The error correction method is used to improve the prediction accuracy.;

Financial markets play an essential role in developing modern society and enabling the deployment of economic resources. This study focuses on predicting stock prices using deep learning models. In particular, the daily closing prices of two different stocks from the Casablanca Stock Market Viz Bank of Africa and Itissalat Al-Maghrib (IAM) are considered. The datasets were pre-processed and passed through the Long Short-Term Memory (LSTM), Multi-Layer Perceptron (MLP), and Convolutional Neural Networks (CNN) models. The models’ performances were compared based on the performance evaluation metrics, viz: mean squared error (MSE) and root mean squared error (RMSE) and Mean Absolute Error (MAE). The paper proposes a novel hybrid model. The hybrid design of the model improves its predictive power as the results of the Hybrid network performance surpassed all the other models.

Improving forecasts, particularly the accuracy, efficiency, and precision of time-series forecasts, is becoming critical for authorities to predict, monitor, and prevent the spread of the Coronavirus disease. However, the results obtained from the predictive models are imprecise and inefficient because the dataset contains linear and non-linear patterns, respectively. Linear models such as autoregressive integrated moving average cannot be used effectively to predict complex time series, so nonlinear approaches are better suited for such a purpose. Therefore, to achieve a more accurate and efficient predictive value of COVID-19 that is closer to the true value of COVID-19, a hybrid approach was implemented. Therefore, the objectives of this study are twofold. The first objective is to propose intelligence-based prediction methods to achieve better prediction results called autoregressive integrated moving average–least-squares support vector machine. The second objective is to investigate the performance of these proposed models by comparing them with the autoregressive integrated moving average, support vector machine, least-squares support vector machine, and autoregressive integrated moving average–support vector machine. Our investigation is based on three COVID-19 real datasets, i.e., daily new cases data, daily new death cases data, and daily new recovered cases data. Then, statistical measures such as mean square error, root mean square error, mean absolute error, and mean absolute percentage error were performed to verify that the proposed models are better than the autoregressive integrated moving average, support vector machine model, least-squares support vector machine, and autoregressive integrated moving average–support vector machine. Empirical results using three recent datasets of known the Coronavirus Disease-19 cases in Malaysia show that the proposed model generates the smallest mean square error, root mean square error, mean absolute error, and mean absolute percentage error values for training and testing datasets compared to the autoregressive integrated moving average, support vector machine, least-squares support vector machine, and autoregressive integrated moving average–support vector machine models. This means that the predicted value of the proposed model is closer to the true value. These results demonstrate that the proposed model can generate estimates more accurately and efficiently. Compared to the autoregressive integrated moving average, support vector machine, least-squares support vector machine, and autoregressive integrated moving average–support vector machine models, our proposed models perform much better in terms of percent error reduction for both training and testing all datasets. Therefore, the proposed model is possibly the most efficient and effective way to improve prediction for future pandemic performance with a higher level of accuracy and efficiency. [ABSTRACT FROM AUTHOR]

Drought is a prolonged dry period that has a serious impact on health, agriculture, economies, energy, and the environment. Thus, there have been numerous attempts to make this phenomenon more predictable for preventing the aforementioned effects. The present study aims to determine the best combination of input data sets and predict the Standardized Precipitation Evapotranspiration Index (SPEI) in 1,6, and 12-month time scales using Artificial Intelligence (AI) models (Multilayer Perceptron Neural Network (MLPNN), Support Vector Regression (SVR), Adaptive Neuro-Fuzzy Inference System (ANFIS), and Ensemble Decision Tree (EDT)), which all models are hybridized with a wavelet transformation at three synoptic stations named Ardebil, Khalkhal, and Moghan. To this end, monthly lags of precipitation, temperature, and SPEI were used in northwestern Iran from 1987 to 2018. The methods were classified into single parameter and multiparameter, and each sub-method was designed based on a combination of the parameters. Moreover, Autocorrelation Function (ACF), Partial Autocorrelation Function (PACF), Effective Factor Elimination Technique (EFET), and Feature Scaling (FS) were used to determine the best lags of parameters. In this regard, Root Mean Square Error (RMSE), Correlation Coefficient (CC), and Nash–Sutcliffe Efficiency Index (NSE) were used as the statistical criteria to assess AI models, methods, and sub-methods. The results revealed that the sub-method (D) with W-MLPNN in the 1-month and 6-month time scales and the sub-method (D, P, T) with W-SVR in the 12-month time scale were the best models and sub-methods of this area, respectively. Moreover, based on the results, the efficiency of the AI was enhanced in longer time scales (more than the 6th month) and the longer the time scale, the more the number of lags (under the 6th month) in input data is decreased. [ABSTRACT FROM AUTHOR]

In the present study, seasonal autoregressive integrated moving average (SARIMA) time series models, nonlinear BL model, multi-layer perceptron artificial neural network and SARIMA-bilinear hybrid models were employed to predict the quality parameters of total dissolved solids (TDS), sodium adsorption ratio (SAR), electrical conductivity (EC) in the Maroon basin in Khuzestan Province. For fitting the mentioned models, the monthly data were used for the calibration (1970–2000), confirmation (2001–2011) and prediction of the model (2012–2018). The appropriate models of SARIMA, the bilinear BL and SARIMA-BL hybrid models of the rationalized parameters of the above-mentioned quality parameters were selected based on the adequacy tests, such as the Akaike criterion and the independence test of the model residuals (Ljung–Box). To determine the effective input parameters of the network, the partial mutual information (PMI) algorithm was used to model the three monthly EC, TDS, and SAR parameters. Also, for input and output layers linear transfer function and for hidden layer various active functions with the back-propagation learning algorithm were used for modeling and predicting this water quality parameters. Comparison of the models showed the tangible superiority of SARIMA-bilinear hybrid models than the artificial neural network with effective input parameters based on PMI algorithm, SARIMA and bl models in performed prediction of the all three monthly qualitative parameters in all three stages, training-validation-test in Maroon basin. [ABSTRACT FROM AUTHOR]

The Artificial Fish Swarm Algorithm (AFSA) is inspired by the ecological behaviors of fish schooling in nature, viz., the preying, swarming and following behaviors. Owing to a number of salient properties, which include flexibility, fast convergence, and insensitivity to the initial parameter settings, the family of AFSA has emerged as an effective Swarm Intelligence (SI) methodology that has been widely applied to solve real-world optimization problems. Since its introduction in 2002, many improved and hybrid AFSA models have been developed to tackle continuous, binary, and combinatorial optimization problems. This paper aims to present a concise review of the continuous AFSA, encompassing the original ASFA, its improvements and hybrid models, as well as their associated applications. We focus on articles published in high-quality journals since 2013. Our review provides insights into AFSA parameters modifications, procedure and sub-functions. The main reasons for these enhancements and the comparison results with other hybrid methods are discussed. In addition, hybrid, multi-objective and dynamic AFSA models that have been proposed to solve continuous optimization problems are elucidated. We also analyse possible AFSA enhancements and highlight future research directions for advancing AFSA-based models. [ABSTRACT FROM AUTHOR]

Water scarcity is the most obstacle faced by irrigation water requirements, likewise, limited available meteorological data to calculate reference evapotranspiration. Consequently, the focal aims of the investigation are to assess the potential of machine learning models in forecasting irrigation water requirements (IWR) of snap beans by evolving multi-scenarios of inputs parameters to figure out the impact of meteorological, crop, and soil parameters on IWR. Six models were applied, support vector regressor (SVR), random forest (RF), deep neural networks (DNN), convolutional neural networks (CNN), long short-term memory (LSTM), and Hybrid CNN-LSTM. Ten variables including maximum and minimum temperature, Relative humidity, wind speed, precipitation, root depth, basal crop coefficient, soil evaporation, a fraction of surface wetted and, exposed and soil wetted fraction were used as the input data for models with their combination, 8 input scenarios were designed. Overall models, the best scenario was scenario 4 (relative humidity, wind speed, basal crop coefficient, soil evaporation), however, the best scenario for DNN and RF model was scenario 7 (root depth, basal crop coefficient, soil evaporation, fraction of surface wetted, exposed and soil wetted fraction). While the weakest one was the group of climatic factors in scenario 6 (maximum temperature, minimum temperature, relative humidity, wind speed, and precipitation). Among the models, the hybrid LTSM & CNN was the most accurate and the SVR model had the lowest estimation accuracy. The outcomes of this research work could set up a modeling strategy that would set in motion the improvement of efforts to identify the shortages in IWR forecasting, which sequentially may support alleviation strategies such as policies for sustainable water use and water resources management. The current approach was promising and has research value for other similar regions. [ABSTRACT FROM AUTHOR]

Soil temperature (ST) is one of the vital parameters of soil, which can affect many of the processes occurring in soil. Accurate techniques are therefore needed to know the ST regime at different soil depths. The present study strived to model the daily ST of the Isfahan and Rasht stations, Iran, at four multiple depths (5, 10, 50, 100 cm). To achieve this goal, the standalone gene expression programming (GEP) and support vector machine (SVM) models were applied. In this context, two types of input patterns were taken into consideration including weather parameters- and lagged ST data-based patterns. Additionally, two different pre-processing methods comprising of entropy and τ-Kendall were utilized to discern the most important weather parameters under the weather parameters-based patterns. It was found that both the employed pre-processing techniques introduced various input combinations as the most important weather parameters. In general, entropy-based GEP and SVM models slightly presented higher accuracy compared to the τ-Kendall-based standalone models at different depths. Then, wavelet (W)-based coupled models were proposed by hybridizing the W theory on the GEP and SVM in order to develop the hybrid W-GEP and W-SVM approaches. The outcomes demonstrated that the implemented coupled models outperformed the single GEP and SVM in modeling the ST of various depths under both the defined input patterns. [ABSTRACT FROM AUTHOR]

Quick population increase and the desire for urbanization are the main drivers for accelerating urban expansion on agricultural lands in Egypt. This issue is obvious in governorates with no desert backyards. This study aims to (1) explore the trend of Land Use Land Cover Change (LULCC) through the period of 1991–2018; (2) upgrade the reliability of predicting LULCC by integrating the Cellular Automata (CA)-Markov chain and fuzzy analytical hierarchy process (FAHP); and (3) perform analysis of urbanization risk on LST trends over the Gharbia governorate for the decision makers to implement effective strategies for sustainable land use. Multi-temporal Landsat images were used to monitor LULCC dynamics from 1991 to 2018 and then simulate LULCC in 2033 and 2048. Two comparable models were adopted for the simulation of spatiotemporal dynamics of land use in the study area: CA-Markov chain and FAHP-CA-Markov chain hybrid models. The second model upgrades the potential of the CA-Markov chain for prediction by its integration with FAHP, which can determine the locations of high potential to be urbanized. The outcomes stated a significant LULCC in Gharbia during the study period—specifically, urban sprawl on agricultural land, and this trend is predicted to carry on. The agricultural sector represented 91.2% in 1991 and reduced to 83.7% in 2018. The built-up area is almost doubled by 2048 with respect to 2018. The regression analysis revealed the LST increase due to urbanization, causing an urban heat island phenomenon. Criteria-based analysis reveals the district's vulnerability to rapid urbanization, which is efficient for data-gap zones. The simulation results make sense since the FAHP-CA-Markov simulated the LULCC in a thoughtful way, considering the driving forces of LULCC, while the CA-Markov chain results were relatively random. Therefore, the FAHP-CA-Markov chain is the pioneer to be relied upon for future projection. The findings of this work provide a better understanding of LULCC trends over the years supporting decision makers toward sustainable land use. Thus, further urbanization should be planned to avert the loss of agricultural land and uninterrupted increasing temperatures. [ABSTRACT FROM AUTHOR]

Groundwater level fluctuations are one of the main components of the hydrogeological cycle and one of the required variables for many water resources operation models. The numerical models can estimate groundwater level (GWL) based on extensive statistics and information and using complex equations in any area. But one of the most important challenges in analyzing and predicting groundwater depletion in water management is the lack of reliable and complete data. For this reason, the use of artificial intelligence models with high predictive accuracy and due to the need for less data is inevitable. In recent years, the use of different numerical models has been noticed as an efficient solution. These models are able to estimate groundwater levels in any region based on extensive statistics and information and also various field experiments such as pumping tests, geophysics, soil and land use maps, topography and slope data, different boundary conditions and complex equations. In the current research, first, by using available statistics, information and maps, the groundwater level fluctuations of the Sonqor plain are simulated by the GMS model, and the accuracy of the model is evaluated in two stages of calibration and validation. Then, due to the need for much less data volume in artificial intelligence-based methods, the GA-ANN and ICA-ANN hybrid methods and the ELM and ORELM models are utilized. The results display that the output of the ORELM model has the best fit with observed data with a correlation coefficient equal to 0.96, and it also has the best and closest scatter points around the 45 degrees line, and in this sense, it is considered as the most accurate model. To ensure the correct selection of the best model, the Taylor diagram is also used. The results demonstrate that the closest point to the reference point is related to the ORELM method. Therefore, to predict the groundwater level in the whole plain, instead of using the complex GMS model with a very large volume of data and also the very time-consuming process of calibration and verification, the ORELM model can be used with confidence. This approach greatly helps researchers to predict groundwater level variations in dry and wet years using artificial intelligence with high accuracy instead of numerical models with complex and time-consuming structures. [ABSTRACT FROM AUTHOR]

Achieving the desired accuracy in time series forecasting has become a binding domain, and developing a forecasting framework with a high degree of accuracy is one of the most challenging tasks in this area. Combining different forecasting methods to construct efficient hybrid models has been widely reported in the literature regarding this challenge. Various types of hybrid models have been developed and successfully employed to improve forecasting accuracy. The well-known hybrid models can be generally categorized into four classes: (1) preprocessing-based, (2) parameter optimization-based, (3) components combination-based, and (4) postprocessing-based hybrid models. Despite the significant successes of hybrid models, efforts to access more accurate results face continued growth. Hybridization of hybrid models is a novel idea proposed to obtain extreme accuracy in recent literature, in which two or more hybrid classes are combined instead of conjoining the conventional individual forecasting methods. Although, in many publications, the aforementioned classes of hybrid models have been reviewed and analyzed in a wide variety of forecasting fields; no study is conducted to review the hybridization of hybrid models. This paper's main contribution is to fill this gap and provide classification and comprehensive review of the current endeavors done in the hybridization of hybrid models in time series forecasting areas. Our searches indicate that more than 250 papers have been published in recent years utilizing hybridization of hybrid models. In this paper, these published papers have been classified regarding their different used combination strategies into four main categories, including (1) Hybridization with preprocessing-based hybrid models (HPH), (2) Hybridization with parameter optimization-based hybrid models (HOH), (3) Hybridization with components combination-based hybrid models (HCH) and, (4) Hybridization with postprocessing-based hybrid models (HSH). Each hybridization of the hybrid class is evaluated regarding the usage frequency, specific merits, and limitations. It can be inferred from reviewing articles that the hybridization of the hybrid concept, as a recent advancement in time series forecasting, can significantly improve traditional hybrid models' accuracy. Furthermore, each category's research gaps and some future research directions are identified in this paper. [ABSTRACT FROM AUTHOR]

Antarctic ecosystems are under increasing anthropogenic pressure, but efforts to predict the responses of Antarctic biodiversity to environmental change are hindered by considerable data challenges. Here, we illustrate how novel data capture technologies provide exciting opportunities to sample Antarctic biodiversity at wider spatiotemporal scales. Data integration frameworks, such as point process and hierarchical models, can mitigate weaknesses in individual data sets, improving confidence in their predictions. Increasing process knowledge in models is imperative to achieving improved forecasts of Antarctic biodiversity, which can be attained for data-limited species using hybrid modelling frameworks. Leveraging these state-of-the-art tools will help to overcome many of the data scarcity challenges presented by the remoteness of Antarctica, enabling more robust forecasts both near- and long-term. Antarctic ecosystems are important contributors to global climatic and biogeochemical cycles, and represent simplified global sentinels that can provide insights into how biodiversity will respond to global change. New techniques to capture biodiversity data offer promising solutions to overcome the considerable logistical constraints faced by biologists in Antarctica. Advancements in data and model integration allow the combination of limited biodiversity data sets to improve data coverage and, therefore, forecasts of Antarctic ecosystems. Leveraging available data, however sparse, while increasingly focusing on incorporating ecological processes into flexible, hybrid modelling frameworks can increase the capability and accuracy of biodiversity forecasts in Antarctica. [ABSTRACT FROM AUTHOR]

The amount of oxygen blown into the converter directly affects the productivity of the converter and the quality of the molten steel, so it is crucial to accurately predict oxygen consumption. Herein, a hybrid model for predicting oxygen consumption in basic oxygen furnace steelmaking process is established. The oxygen consumption of converter is divided into two types: the determined oxygen consumption and the estimated oxygen consumption. The determined oxygen consumption is calculated by mechanism model. And the estimated oxygen consumption is calculated by statistical model. The method of cluster analysis is also introduced into the model. After clustering the data, a statistical model for predicting the estimated oxygen consumption in each subcategory is established, which further improves the prediction accuracy of the model. After dividing the original data into three categories, the hit rate of the hybrid model is 93.3%, when the relative errors of the model are restricted within 5%. As a result, the hybrid model is fully capable of predicting oxygen consumption in converters. [ABSTRACT FROM AUTHOR]

Many of the deep learning solutions for time-series forecasting reported in the literature include complex neural networks that may not be directly employed by the practitioner in the field. In this study, we demonstrate how the standard deep neural network types, convolutional neural network (CNN) and long short-term memory (LSTM) network can be applied in the field of time-series forecasting. This study consists of two parts. The first part is to compare CNN and LSTM models with classical methods like Random Forest (RF) and ARIMA for the univariate electric power consumption task. The second part is to use the best performing model from the first part in the hybrid model and perform data fusion with the newly built hybrid model for the electric power consumption forecasting task. CNN and LSTM models outperform traditional methods when their performances are evaluated on the univariate electric power consumption data of Illinois, USA. We also illustrate the use of hybrid deep learning models composed of standard CNN and LSTM for data fusion with the aim of time-series forecasting. When the hybrid models are applied to the fused data of the electric power consumption data and the multivariate weather data of Illinois, USA, the forecasting performance is improved compared to that when only univariate data is used. [ABSTRACT FROM AUTHOR]

Abstract The development of sequence decomposition techniques in recent years has facilitated the wide use of decomposition-based prediction models in hydrological forecasting. However, decomposition-based prediction models usually use the overall decomposition (OD) sampling technique to extract samples. Some studies have shown that the OD sampling technique causes abnormally “high” performance of models owing to the utilization of future information, and this technique cannot be applied in practice. Several researchers have also proposed novel sampling techniques, such as semi-stepwise decomposition (SSD), fully stepwise decomposition (FSD), and single-model SSD (SMSSD). Moreover, an improved single-model FSD (SMFSD) sampling technique is proposed in this study. Four decomposition methods, namely discrete wavelet transform (DWT), empirical mode decomposition (EMD), complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), and variational mode decomposition (VMD), are introduced in this study. A systematic investigation of the models developed using OD sampling techniques is conducted, and the applicability of SSD, FSD, SMSSD, and SMFSD sampling techniques is reasonably evaluated. The application of monthly runoff prediction using the five sampling techniques and four decomposition methods at five representative hydrological stations in Poyang Lake, China, shows that (1) EMD and CEEMDAN (including the improved EMD-based adaptive decomposition method) cannot be used to construct stepwise decomposition prediction models because the implementation of the stepwise decomposition strategy leads to a variable number of sub-series. (2) OD sampling techniques cannot develop convincing models for practical prediction because future information is introduced into the samples for model training. (3) Models developed based on SSD and SMSSD sampling techniques do not use future information in the training process, but suffer from severe overfitting and inferior prediction performance. (4) Models developed based on FSD and SMFSD sampling techniques can produce convincing prediction results, and the combination of the proposed SMFSD sampling technique and VMD develops prediction models with superior performance and significantly enhances the efficiency of the models.

Digital Twins (DTs) are on the rise as innovative, powerful technologies to harness the power of digitalisation in the WRRF sector. The lack of consensus and understanding when it comes to the definition, perceived benefits and technological needs of DTs is hampering their widespread development and application. Transitioning from traditional WRRF modelling practice into DT applications raises a number of important questions: When is a model's predictive power acceptable for a DT? Which modelling frameworks are most suited for DT applications? Which data structures are needed to efficiently feed data to a DT? How do we keep the DT up to date and relevant? Who will be the main users of DTs and how to get them involved? How do DTs push the water sector to evolve? This paper provides an overview of the state-of-the-art, challenges, good practices, development needs and transformative capacity of DTs for WRRF applications. HIGHLIGHTS A Digital Twin distinguishes itself from a simulation model by a continuous, automated data connection.; Current DT projects in WRRFs focus on operational support and control.; Combining mechanistic models and data-driven techniques into hybrid models can accelerate the adoption of DTs.; Data and information models are key for the implementation and upscaling of DTs.; WRRF staff should be included during the development stages of DTs.;

Epilepsy is a complex neurological disorder marked by recurrent seizures, often stemming from abnormal discharge of the brain. Electroencephalogram (EEG) captures temporal and spatial shifts in cerebral electrical activity, holding pivotal diagnostic and therapeutic value for epilepsy. Deep learning techniques have made remarkable progress in EEG-based seizure detection over recent years. This review is dedicated to exploring seizure detection approaches based on deep learning, focusing on three distinct avenues. Primarily, we delve into the application of canonical deep learning methods in epilepsy detection. Subsequently, a more in-depth study was conducted on the hybrid models of deep learning. Next, the third is the integration of deep learning and traditional machine learning strategies. Finally, the challenges and future prospects related to this topic are put forward. The uniqueness of this review lies in its novel and comprehensive perspective on the latest research on deep learning-based epilepsy detection by systematically classifying methods, visualizing research progress, and addressing challenges and gaps in current research. It can provide valuable guidance for researchers who want to delve into the field of epileptic seizure detection based on EEG signals. • This paper provides a novel and comprehensive perspective on the latest research on deep learning-based seizure detection. • Three methods based on deep learning are described systematically: Canonical DL Methods, Hybrid DL Methods and Integrating DL with Traditional ML Methods. • The existing challenges and the prospects of further research on epilepsy detection by deep learning methods are summarized. [ABSTRACT FROM AUTHOR]

Vision Transformers (ViTs) have achieved significant advancement in computer vision tasks due to their powerful modeling capacity. However, their performance notably degrades when trained with insufficient data due to lack of inherent inductive biases. Distilling knowledge and inductive biases from a Convolutional Neural Network (CNN) teacher has emerged as an effective strategy for enhancing the generalization of ViTs on limited datasets. Previous approaches to Knowledge Distillation (KD) have pursued two primary paths: some focused solely on distilling the logit distribution from CNN teacher to ViT student, neglecting the rich semantic information present in intermediate features due to the structural differences between them. Others integrated feature distillation along with logit distillation, yet this introduced alignment operations that limits the amount of knowledge transferred due to mismatched architectures and increased the computational overhead. To this end, this paper presents H ybrid D ata-efficient K nowledge D istillation (HDKD) paradigm which employs a CNN teacher and a hybrid student. The choice of hybrid student serves two main aspects. First, it leverages the strengths of both convolutions and transformers while sharing the convolutional structure with the teacher model. Second, this shared structure enables the direct application of feature distillation without any information loss or additional computational overhead. Additionally, we propose an efficient light-weight convolutional block named M o b ile C hannel- S patial A ttention (MBCSA), which serves as the primary convolutional block in both teacher and student models. Extensive experiments on two medical public datasets showcase the superiority of HDKD over other state-of-the-art models and its computational efficiency. Source code at: https://github.com/omarsherif200/HDKD. [ABSTRACT FROM AUTHOR]

Digital technologies with predictive modelling capabilities are revolutionizing electricity markets, especially in demand-side management. Accurate electricity price prediction is essential in deregulated markets; however, developing effective models is challenging due to high-frequency fluctuations and price volatility. This study introduces a hybrid prediction system that addresses these challenges through a comprehensive data processing and modelling framework for half-hourly electricity price predictions. The preprocessing stage employs the Maximum Overlap Discrete Wavelet Transform (MoDWT) to enhance input quality by reducing overlap and revealing underlying price patterns. The prediction model integrates Convolutional Neural Networks with Random Vector Functional Link (CRVFL) in a deep learning hybrid approach. Bayesian Optimization fine-tunes the MoDWT-CRVFL model for optimal performance. Validation of the model is conducted using half-hourly electricity prices from New South Wales. The results highlight the efficacy of the MoDWT-CRVFL model, achieving high accuracy with superior Global Performance Indicator (G P I) values of approximately 1.61, 1.33, 1.85, 1.30, and 0.78 for Summer, Autumn, Winter, Spring, and Annual (Year 2022), respectively, outperforming alternative models. Similarly, the Kling–Gupta Efficiency (K G E) metrics for the proposed model consistently surpassed those of both decomposition-based and standalone models. For instance, the K G E value for MoDWT-CRVFL was approximately 0.972, significantly higher than values of approximately 0.958, 0.899, 0.963, 0.943, 0.930, 0.661, 0.708, 0.696, 0.739, and 0.738 for MoDWT-LSTM, MoDWT-DNN, MoDWT-XGB, MoDWT-RF, MoDWT-MLP, Bi-LSTM, LSTM, DNN, RF, XGB, and MLP, respectively. The methodologies proposed in this study optimize energy resource allocation, market prices, and network management, empowering market operators to make informed decisions for a resilient and efficient electricity market. • Designs deep hybrid Convolutional Neural-Random Vector Functional Link-Wavelet data decomposition model. • Performs half-hourly electricity price predictions over seasonal and annual periods. • Performs detailed appraisal of proposed deep hybrid decomposition model with visual and statistical predictive analytics. • Proposed deep hybrid decomposition model outperforms data decomposition and standalone predictive models. • Demonstrates a practical method for seasonal/annual electricity market insight analysis. [ABSTRACT FROM AUTHOR]

• Hybrid deep learning model outperforms basic LSTM in Bitcoin forecasting. • NRBO-CNN-BiLSTM-Attention model cuts MAPE by over 50 %. • Significant accuracy in 5-day and 15-day Bitcoin price forecasts. • Model tested across multiple loss functions for robustness. • Suggests dataset restructuring to improve model performance. This paper employs a deep learning network with a comprehensive architecture to forecast Bitcoin prices, enhancing accuracy by integrating two meta-heuristic optimization algorithms, INFO and NRBO. Empirical results demonstrate that the hybrid model significantly outperforms the LSTM in both fit and predictive accuracy across in-sample and out-of-sample data. Notably, the NRBO-CNN-BiLSTM-Attention model substantially improves accuracy in 5-day and 15-day forecasts, reducing the MAPE by over 50 % compared to the LSTM model, thereby significantly enhancing overall predictive performance. The robustness of our results is supported by the MCS tests. Furthermore, strategically modifying time steps in data analysis optimizes model performance. [ABSTRACT FROM AUTHOR]

• Hybrid models outperformed individual models on a national scale. • Sentinel-2 data showed significant potential for predicting SOC. • The RFK model at 100m resolution gave the best prediction in Germany. • Spatial resolutions affected hybrid models' performance. • The predicted map provided the spatial variation of SOC in Germany. Mapping the spatial distribution of soil organic carbon (SOC) is crucial for monitoring soil health, understanding ecosystem functions, and contributing to global carbon cycling. However, few studies have directly compared the influence of hybrid models and individual models with varying spatial resolutions on SOC prediction at a national scale. In this study, by combining remote sensing data, we utilized the LUCAS 2018 soil dataset to evaluate the potential capacities of hybrid models for predicting SOC content at different spatial resolutions in Germany. The hybrid models PLSRK and RFK consisted of partial least square regression (PLSR) with residual original kriging (OK) models, and random forest (RF) models with residual OK models, respectively. Individual PLSR and RF models were used as reference models. All these models were applied to estimate SOC content at 10 m, 50 m, 100 m, and 200 m spatial resolutions. Sentinel-2 bands, band indices, and topography variables were as predictors. The results revealed that hybrid models had a more accurate prediction of SOC content with higher explanations and lower prediction errors compared with individual models. The RFK model at the spatial resolution of 100 m was the fittest model with R 2 = 0.416, RMSE = 0.545, and RPIQ = 1.647, which enhanced 3.74% of explanation compared with the performance of RF model. The results also showed that hybrid models at a relatively coarse resolution (100 m) had better accuracy instead of those at high spatial resolution (10 m, 50 m). Sentinel-2 remote sensing data showed significant predictive capabilities for estimating SOC content. The predicted spatial distribution of SOC content revealed that the high SOC concentrated in the northwest grassland, central and southwestern mountains, and the Alps in Germany. Our study provided a benchmark SOC map in Germany for monitoring the changes resulting from land use and climate impacts, and we illustrated the accuracy of hybrid models and the effects of spatial resolutions on SOC predictions at a national scale. [ABSTRACT FROM AUTHOR]

Forecasting of electricity consumption is a critical issue, due to its importance in the planning of the energy trading countries. Several new techniques such as hybrid models are used as well as classical single models to estimate electricity consumption. This study aims to get the best electricity consumption model of Türkiye. For this, the forecasting performances of single and hybrid electricity consumption models, SARIMA is the time series model, ANNs and MLPs are machine learning single models and SARIMA-ANNs and SARIMA-MLPs are hybrid models of machine learning, are compared. This study employs new hybrid models and examines whether the multiplicative model of Wang et al. or the combined model of Khashei and Bijari is superior to than Zhang's hybrid model commonly used as the ARIMA-hybrid model with well known flaws. The results show that hybrid models are more accurate than single time series/machine learning models when forecasting Turkish electricity consumption. Moreover, The Khashei and Bijari hybrid model outperformed the other models and it was determined as the best model for forecasting Türkiye's electricity consumption. • In addition to single time series and machine learning models, three types of hybrid models were examined: Zhang, Wang et al. and Khashei and Bijari. • RMSE, MAE, R-squared criteria and two types of Diebold-Mariano test are used for model comparison. • The hybrid models are more accurate than single time series/machine learning models. • ANN GRprop algorithm is a more efficient algorithm than the back-propagation. • The Khashei and Bijari is the most successful hybrid model for forecasting the Türkiye's electricity consumption. [ABSTRACT FROM AUTHOR]

Soil temperature patterns are of great importance for any agro-based economy like Bangladesh since they significantly affect biological, chemical, and physical processes that take place in the soil. Unfortunately, there have been no forecast studies on soil temperature in Bangladesh until now. In this article, we used five tree-based models (decision tree, random forest, gradient boosting tree, a hybrid of decision tree and gradient boosting tree, and a hybrid of random forest and gradient boosting tree) to mine strong links among different meteorological factors and soil temperature at different time window sizes. We found that a hybrid of random forest and gradient boosting tree with all the meteorological factors and a five-day time window is optimal for forecasting soil temperature at depths of 10 cm and 30 cm for all lead times (one, three, or five days), whereas the random forest with the same input scenario and time window is optimal for forecasting soil temperature at a depth of 50 cm for long lead times (five days). Since our study includes the first soil temperature forecast model in Bangladesh, it provides valuable insights for agricultural soil management, fertilizer application, and water resource optimization in Bangladesh, as well as in other South Asian countries that share the same climate patterns as Bangladesh.