Hybrid Support Vector Regression Research Articles

Hierarchical parameter optimization based support vector regression for power load forecasting

Power load forecasting is an important task of smart grid, which is of great significance to the sustainable development of society. In this paper, a hybrid support vector regression (HSVR) is raised for the medium and long term load forecasting. To further improve prediction accuracy, the coupling and interdependent relationship between hyperparameters and model parameters in the optimization process is focused. A hierarchical optimization method based on nested strategy and state transition algorithm (STA) is proposed to find optimal parameters. The effectiveness of the proposed hierarchical optimization method is confirmed on several benchmarks, and the resulting hierarchical optimization method based SVR is also successfully applied to a real industrial power load forecasting problem in China.

Prediction of meteorological drought by using hybrid support vector regression optimized with HHO versus PSO algorithms.

Drought is considered one of the costliest natural disasters that result in water scarcity and crop damage almost every year. Drought monitoring and forecasting are essential for the efficient management of water resources and sustainability in agriculture. However, the design of a consistent drought prediction model based on the dynamic relationship of the drought index with its antecedent values remains a challenging task. In the present research, the SVR (support vector regression) model was hybridized with two different optimization algorithms namely; Particle Swarm Optimization (PSO) and Harris Hawks Optimization (HHO) for reliable prediction of effective drought index (EDI) 1 month ahead, at different locations of Uttarakhand State of India. The inputs of the models were selected through partial autocorrelation function (PACF) analysis. The output produced by the SVR-HHO and SVR-PSO models was compared with the EDI estimated from observed data using five statistical indicators, i.e., RMSE (Root Mean Square Error), MAE (Mean Absolute Error), COC (Coefficient of Correlation), NSE (Nash-Sutcliffe Efficiency), WI (Willmott Index), and graphical inspection of radar-chart, time-variation plot, box-whisker plot, and Taylor diagram. Appraisal of results indicates that the SVR-HHO model (RMSE = 0.535-0.965, MAE = 0.363-0.622, NSE = 0.558-0.860, COC = 0.760-0.930, and WI = 0.862-0.959) outperformed the SVR-PSO model (RMSE = 0.546-0.967, MAE = 0.372-0.625, NSE = 0.556-0.855, COC = 0.758-0.929, and WI = 0.861-0.956) in predicting EDI. Visual inspection of model performances also showed a better performance of SVR-HHO compared to SVR-PSO in replicating the median, inter-quartile range, spread, and pattern of the EDI estimated from observed rainfall. The results indicate that the hybrid SVR-HHO approach can be utilized for reliable EDI predictions in the study area.

Support vector regression integrated with novel meta-heuristic algorithms for meteorological drought prediction

Drought is a complex natural phenomenon, so, precise prediction of drought is an effective mitigation tool for measuring the negative consequences on agriculture, ecosystems, hydrology, and water resources. The purpose of this research was to explore the potential capability of support vector regression (SVR) integrated with two meta-heuristic algorithms i.e., Grey Wolf Optimizer (GWO), and Spotted Hyena Optimizer (SHO), for meteorological drought (MD) prediction by utilizing EDI (effective drought index). For this objective, the two-hybrid SVR–GWO, and SVR–SHO models were constructed at Kumaon and Garhwal regions of Uttarakhand State (India). The EDI was computed in both study regions by using monthly rainfall data series to calibrate and validate the advanced hybrid SVR models. The autocorrelation function (ACF) and partial-ACF (PACF) were utilized to determine the optimal inputs (antecedent EDI) for EDI prediction. The results produced by the hybrid SVR models were compared with the calculated (observed) values by employing the statistical indicators and through graphical inspection. A comparison of results demonstrates that the hybrid SVR–GWO model outperformed to the SVR–SHO models for all study stations located in Kumaon and Garhwal regions. Also, the results highlighted the better suitability, supremacy, and convergence behavior of meta-heuristic algorithms (i.e., GWO and SHO) for meteorological drought prediction in the study regions.

A hybrid support vector regression with multi-domain features for low-velocity impact localization on composite plate structure

The accurate localization of low-velocity impacts on the composite plate structure of the ship is still a great challenge. Current research mainly focuses on extracting single domain features from impact signals as the input of machine learning methods, whereas ignores multi-domain features with more comprehensive impact information. In this paper, a hybrid support vector regression with multi-domain features is proposed to increase the localization accuracy in determining the locations of low-velocity impacts on the composite plate structure. The proposed method consists of the signal preprocessing, the multi-domain feature extraction, and the impact localization. In the signal preprocessing, the trend component in the low-velocity impact signals is eliminated by adopting the empirical mode decomposition (EMD) method. Then, the multi-domain features, which include time domain features, frequency domain features, and time-frequency domain features, are extracted from the preprocessed impact signals. Finally, the optimized support vector regression based on the bat algorithm (BA-SVR) is designed to implement the localization of low-velocity impacts. The low-velocity impact localization system using four fiber Bragg grating (FBG) sensors is established on a carbon fiber reinforced plastic (CFRP) plate, and then five sets of experiments are executed. The statistical results in these experiments demonstrate the effectiveness and feasibility of BA-SVR that uses multi-domain features and four FBG sensors and the satisfactory localization performance of the proposed method in handling the low-velocity impact localization problem on the CFRP plate.

Forecasting Framework Using Hybrid Modeling and Support Vector Regression

Time series forecasting for financial market has increasingly attracted the interests of investors and academic researchers. In recent years, some hybrid models have been constructed to improve the predictions since some methods cannot extract useful information from stock time series with noise to conduct prediction. In this research, a prediction framework is proposed to forecast the stock market behavior using the methods of wavelet coherence, multiscale decomposition and support vector regression (SVR). First, a combined method is applied to the raw data and remove noise to get useful information. Then, a SVR model is applied to improve the prediction performance of multidimensional nonlinear data. Furthermore, the comparison experiments were performed with both Shanghai Composite Index and Dow Jones Index to examine the effectiveness of the framework. The results indicate that the proposed framework performs better than other advanced models.

Estimation of Daily Suspended Sediment Load Using a Novel Hybrid Support Vector Regression Model Incorporated with Observer‐Teacher‐Learner‐Based Optimization Method

Predicting suspended sediment load (SSL) in water resource management requires efficient and reliable predicted models. This study considers the support vector regression (SVR) method to predict daily suspended sediment load. Since the SVR has unknown parameters, the observer‐teacher‐learner‐based Optimization (OTLBO) method is integrated with the SVR model to provide a novel hybrid predictive model. The SVR combined with the genetic algorithm (SVR‐GA) is used as an alternative model. To explore the performance and application of the proposed models, five input combinations of rainfall and discharge data of Cham Siah River catchment are provided. The predictive models are assessed using various numerical and visual indicators. The results indicate that the SVR‐OTLBO model offers a higher prediction performance than other models employed in the current study. Specifically, SVR‐OTLBO model offers highest Pearson correlation coefficient (R = 0.9768), Willmott’s Index (WI = 0.9812), ratio of performance to IQ (RPIQ = 0.9201), and modified index of agreement (md = 0.7411) and the lowest relative root mean square error (RRMSE = 0.5371) in comparison with SVR‐GA (R = 0.9704, WI = 0.9794, RPIQ = 0.8521, and md = 0.7323, 0.5617) and SVR (R = 0.9501, WI = 0.9734, RPIQ = 0.3229, md = 0.4338, and RRMSE = 1.0829) models, respectively.

Daily pan-evaporation estimation in different agro-climatic zones using novel hybrid support vector regression optimized by Salp swarm algorithm in conjunction with gamma test

Ensuring accurate estimation of evaporation is weighty for effective planning and judicious management of available water resources for agricultural practices. Thus, this work enhances the potential of support vector regression (SVR) optimized with a novel nature-inspired algorithm, namely, Slap Swarm Algorithm (SVR-SSA) against Whale Optimization Algorithm (SVR-WOA), Multi-Verse Optimizer (SVR-MVO), Spotted Hyena Optimizer (SVR-SHO), Particle Swarm Optimization (SVR-PSO), and Penman model (PM). Daily EP (pan-evaporation) was estimated in two different agro-climatic zones (ACZ) in northern India. The optimal combination of input parameters was extracted by applying the Gamma test (GT). The outcomes of the hybrid of SVR and PM models were equated with recorded daily EP observations based on goodness-of-fit measures along with graphical scrutiny. The results of the appraisal showed that the novel hybrid SVR-SSA-5 model performed superior (MAE = 0.697, 1.556, 0.858 mm/day; RMSE = 1.116, 2.114, 1.202 mm/day; IOS = 0.250, 0.350, 0.303; NSE = 0.0.861, 0.750, 0.834; PCC = 0.929, 0.868, 0.918; IOA = 0.960, 0.925, 0.956) than other models in testing phase at Hisar, Bathinda, and Ludhiana stations, respectively. In conclusion, the hybrid SVR-SSA model was identified as more suitable, robust, and reliable than the other models for daily EP estimation in two different ACZ.

Prediction of Water Inrush in Long-Lasting Shutdown Karst Tunnels Based on the HGWO-SVR Model

It is very important to predict the long-term shutdown karst tunnel water inrush for preventing tunnel construction accidents. However, it is urgent to study a new prediction model to solve the problems of insufficient sample size and low prediction accuracy for long-term shutdown karst tunnel water inrush prediction. In this study, the water inrush and atmospheric rainfall in a tunnel project in China were monitored for over five months. By adopting hybrid grey wolf optimization (HGWO) algorithm and support vector regression (SVR) method, the HGWO-SVR tunnel water inrush prediction model was proposed. The atmospheric rainfall of the day and yesterday and yesterday’s water inrush were considered in the HGWO-SVR model, and the model was used to predict the tunnel water inrush. The results show that the predicted water inrush value is basically consistent with the measured value. After the parameters of SVR model are optimized by HGWO algorithm, the HGWO-SVR prediction model has the advantages of high precision and less sample demand. The model is more suitable for the prediction of long-term shutdown tunnel water inrush with less measured sample. Thus, the proposed prediction model can effectively be used as a new approach for tunnel water inrush in some similar projects.

Spatial prediction of landslide susceptibility in western Serbia using hybrid support vector regression (SVR) with GWO, BAT and COA algorithms

In this study, we developed multiple hybrid machine-learning models to address parameter optimization limitations and enhance the spatial prediction of landslide susceptibility models. We created a geographic information system database, and our analysis results were used to prepare a landslide inventory map containing 359 landslide events identified from Google Earth, aerial photographs, and other validated sources. A support vector regression (SVR) machine-learning model was used to divide the landslide inventory into training (70%) and testing (30%) datasets. The landslide susceptibility map was produced using 14 causative factors. We applied the established gray wolf optimization (GWO) algorithm, bat algorithm (BA), and cuckoo optimization algorithm (COA) to fine-tune the parameters of the SVR model to improve its predictive accuracy. The resultant hybrid models, SVR-GWO, SVR-BA, and SVR-COA, were validated in terms of the area under curve (AUC) and root mean square error (RMSE). The AUC values for the SVR-GWO (0.733), SVR-BA (0.724), and SVR-COA (0.738) models indicate their good prediction rates for landslide susceptibility modeling. SVR-COA had the greatest accuracy, with an RMSE of 0.21687, and SVR-BA had the least accuracy, with an RMSE of 0.23046. The three optimized hybrid models outperformed the SVR model (AUC = 0.704, RMSE = 0.26689), confirming the ability of metaheuristic algorithms to improve model performance.

A Clustering-Based Hybrid Support Vector Regression Model to Predict Container Volume at Seaport Sanitary Facilities

An accurate prediction of freight volume at the sanitary facilities of seaports is a key factor to improve planning operations and resource allocation. This study proposes a hybrid approach to forecast container volume at the sanitary facilities of a seaport. The methodology consists of a three-step procedure, combining the strengths of linear and non-linear models and the capability of a clustering technique. First, a self-organizing map (SOM) is used to decompose the time series into smaller clusters easier to predict. Second, a seasonal autoregressive integrated moving averages (SARIMA) model is applied in each cluster in order to obtain predicted values and residuals of each cluster. These values are finally used as inputs of a support vector regression (SVR) model together with the historical data of the cluster. The final prediction result integrates the prediction results of each cluster. The experimental results showed that the proposed model provided accurate prediction results and outperforms the rest of the models tested. The proposed model can be used as an automatic decision-making tool by seaport management due to its capacity to plan resources in advance, avoiding congestion and time delays.

Simulation of Titicaca Lake Water Level Fluctuations Using Hybrid Machine Learning Technique Integrated with Grey Wolf Optimizer Algorithm

Lakes have an important role in storing water for drinking, producing hydroelectric power, and environmental, agricultural, and industrial uses. In order to optimize the use of lakes, precise prediction of the lake water level (LWL) is a main issue in water resources management. Due to the existence of nonlinear relations, uncertainty, and characteristics of the time series variables, the exact prediction of the lake water level is difficult. In this study the hybrid support vector regression (SVR) and the grey wolf algorithm (GWO) are used to predict lake water level fluctuations. Also, three types of data preprocessing methods, namely Principal component analysis, Random forest, and Relief algorithm were used for finding the best input variables for prediction LWL by the SVR and SVR-GWO models. Before the LWL simulation on monthly time step using the hybrid model, an evolutionary approach based on different monthly lags was conducted for determining the best mask of the input variables. Results showed that based on the random forest method, the best scenario of the inputs was Xt−1, Xt−2, Xt−3, Xt−4 for the SVR-GWO model. Also, the performance of the SVR-GWO model indicated that it could simulate the LWL with acceptable accuracy (with RMSE = 0.08 m, MAE = 0.06 m, and R2 = 0.96).

Monthly evapotranspiration estimation using optimal climatic parameters: efficacy of hybrid support vector regression integrated with whale optimization algorithm.

For effective planning of irrigation scheduling, water budgeting, crop simulation, and water resources management, the accurate estimation of reference evapotranspiration (ETo) is essential. In the current study, the hybrid support vector regression (SVR) coupled with Whale Optimization Algorithm (SVR-WOA) was employed to estimate the monthly ETo at Algiers and Tlemcen meteorological stations positioned in the north of Algeria under three different optimal input scenarios. Monthly climatic parameters, i.e., solar radiation (Rs), wind speed (Us), relative humidity (RH), and maximum and minimum air temperatures (Tmax and Tmin) of 14years (2000-2013), were obtained from both stations. The accuracy of the hybrid SVR-WOA model was appraised against hybrid SVR-MVO (Multi-Verse Optimizer), and SVR-ALO (Ant Lion Optimizer) models through performance measures, i.e., mean absolute error (MAE), root-mean-square error (RMSE), index of scattering (IOS), index of agreement (IOA), Pearson correlation coefficient (PCC), Nash-Sutcliffe efficiency (NSE), and graphical interpretation (time-variation and scatter plots, radar chart, and Taylor diagram). The results showed that the SVR-WOA model performed superior to the SVR-MVO and SVR-ALO models at both stations in all scenarios. The SVR-WOA-1 model with five inputs (i.e., Tmin, Tmax, RH, Us, Rs: scenario-1) had the lowest value of MAE = 0.0658/0.0489mm/month, RMSE = 0.0808/0.0617mm/month, IOS = 0.0259/0.0165, and the highest value of NSE = 0.9949/0.9989, PCC = 0.9975/0.9995, and IOA = 0.9987/0.9997 for testing period at both stations, respectively. The proposed hybrid SVR-WOA model was found to be more appropriate and efficient in comparison to SVR-MVO and SVR-ALO models for estimating monthly ETo in the study region.

Estimation of monthly and seasonal precipitation: A comparative study using data-driven methods versus hybrid approach

The purpose of this study is to estimate the monthly and seasonal precipitation using historical measured data. Firstly two data-driven methods such as Support Vector Regression (SVR), Gaussian Process Regression (GPR) along a distance-based k-nearest neighbor method (KNN) were used. Secondly, the hybrid Support Vector Regression with the Fruit fly optimization Algorithm (SVR-FOA) was used for comparing the estimated results. The historical measured meteorological data were entered into the model after preprocessing and determination of suitable input combinations based on the feature selection algorithms and sensitivity analysis. The results showed that the hybrid SVR-FOA approach performs better than the other methods, especially during seasonal precipitation estimation. In conclusion, the SVR-FOA model with r of 0.75, NS of 0.56, RMSE of 13.58 mm, and MAE of 10.01 mm during monthly precipitation estimation and r of 0.90, NS of 0.80, RMSE of 6.37 mm and MAE of 5.04 mm during seasonal precipitation is able to accurately estimate precipitation in the study area.

Support vector regression optimized by meta-heuristic algorithms for daily streamflow prediction

Accurate and reliable prediction of streamflow is vital to the optimization of water resources management, reservoir flood operations, catchment, and urban water management. In this research, support vector regression (SVR) was optimized by six meta-heuristic algorithms, namely, Ant Lion Optimization (SVR-ALO), Multi-Verse Optimizer (SVR-MVO), Spotted Hyena Optimizer (SVR-SHO), Harris Hawks Optimization (SVR-HHO), Particle Swarm Optimization (SVR-PSO), and Bayesian Optimization (SVR-BO) to predict daily streamflow in Naula watershed, State of Uttarakhand, India. The significant inputs and parameter combinations for hybrid SVR models were extracted through Gamma Test before processing. The results obtained by hybrid SVR models during calibration (training) and validation (testing) periods, which were compared against observed streamflow using performance indicators of root mean square error (RMSE), scatter index (SI), coefficient of correlation (COC), Willmott index (WI), and by visual inspection (time-series plot, scatter plot and Taylor diagram). The results of comparison demonstrated that SVR-HHO during calibration/validation periods (RMSE = 92.038/181.306 m3/s, SI = 0.401/0.715, COC = 0.881/0.717, and WI = 0.928/0.777) had superior performance to the SVR-ALO, SVR-MVO, SVR-SHO, SVR-PSO, and SVR-BO models in predicting daily streamflow in the study basin. In addition, the new HHO algorithm outperformed the other meta-heuristic algorithms in terms of prediction accuracy.

Short-Term Load Forecasting of Microgrid via Hybrid Support Vector Regression and Long Short-Term Memory Algorithms

Short-Term Load Forecasting (STLF) is the most appropriate type of forecasting for both electricity consumers and generators. In this paper, STLF in a Microgrid (MG) is performed via the hybrid applications of machine learning. The proposed model is a modified Support Vector Regression (SVR) and Long Short-Term Memory (LSTM) called SVR-LSTM. In order to forecast the load, the proposed method is applied to the data related to a rural MG in Africa. Factors influencing the MG load, such as various household types and commercial entities, are selected as input variables and load profiles as target variables. Identifying the behavioral patterns of input variables as well as modeling their behavior in short-term periods of time are the major capabilities of the hybrid SVR-LSTM model. To present the efficiency of the suggested method, the conventional SVR and LSTM models are also applied to the used data. The results of the load forecasts by each network are evaluated using various statistical performance metrics. The obtained results show that the SVR-LSTM model with the highest correlation coefficient, i.e., 0.9901, is able to provide better results than SVR and LSTM, which have the values of 0.9770 and 0.9809, respectively. Finally, the results are compared with the results of other studies in this field, which continued to emphasize the superiority of the SVR-LSTM model.

Video super-resolution using hybrid support vector regression–Actor Critic Neural Network model

The aim of video super-resolution (SR) is to produce a high-resolution (HR) video frame from numerous successive low-resolution (LR) frames. Even though there are a large number of techniques employed for the video SR, all these existing techniques face a hectic challenge at various conditions. Thus, this paper proposes an effective video resolution strategy using the hybrid Support vector regression–Actor Critic Neural Network (SVR–ACNN) model for video enhancement. The SR images formed using the individual SVR model and ACNN are integrated using the weighted average concept. The ACNN is tuned optimally by the proposed Fractional-based Sine Cosine algorithm (F-SCA), which is responsible for the global optimal convergence. The experimentation of the proposed method utilizes three videos taken from the Cambridge-driving Labeled Video Database (CamVid), and the results are analyzed for three scaling factors. The results prove that the proposed model offers a better SR image with a better PSNR, SSIM, and SDME of 33.6447 dB, 0.9398, and 45.2779, respectively.

Spatial prediction of landslide susceptibility using hybrid support vector regression (SVR) and the adaptive neuro-fuzzy inference system (ANFIS) with various metaheuristic algorithms

Landslides are natural and sometimes quasi-natural hazards that are destructive to natural resources and cause loss of human life every year. Hence, preparing susceptibility maps for landslide monitoring is essential to minimizing their negative effects. The main aim of the current research was to develop landslide susceptibility maps for Icheon Township, South Korea, using hybrid Machin learning and metaheuristic algorithms, that is, the bee algorithm (Bee), the adaptive neuro-fuzzy inference system (ANFIS), support vector regression (SVR), and the grey wolf optimizer (GWO), and to compare their predictive accuracy. Based on identified landslide locations, an inventory map was prepared and divided into training and validation data sets (70%/30%). the predicated model outcomes were validated with root mean square error (RMSE), and area under receiver operating characteristic curve (AUC), and pairwise comparison values for the ANFIS, ANFIS–Bee, ANFIS–GWO, SVR, SVR–Bee, and SVR–GWO models were obtained. The area under the curve was obtained with the training and validation data sets. Based on the training data sets, AUC of 80%, 83%, 83%, 69%, 81%, and 80% were obtained for the SVR, SVR–GWO, SVR–Bee, ANFIS, ANFIS–GWO, and ANFIS–Bee models, respectively. For the validation data sets, values of 79%, 82%, 82%, 68%, 79%, and 79%, respectively, were obtained. The SVR–GWO and SVR–Bee models were the most predictive models in terms of constructing the exceptionally focused landslide susceptibility map, with little spatial variation in the highly susceptible classes. Furthermore, the MSE, RMSE, and pairwise comparisons indicated that the SVR–GWO and SVR–Bee models were superior models for this study township. In addition, ANFIS individually was not superior to the ensembles of ANFIS–GWO and ANFIS–Bee for landslide assessment. These landslide susceptibility maps provide a platform for land use planning with an eye toward sustainable development of infrastructure and damage reduction for Icheon Township.

Hybrid SVR-SARIMA model for tourism forecasting using PROMETHEE II as a selection methodology: a Philippine scenario

Purpose This paper aims to propose a hybrid-forecasting model for long-term tourism demand forecasting. As such, it attempts to model the tourism demand in the Philippines, which is a relatively underrepresented area in the literature, despite its tourism sector’s growing economic progress. Design/methodology/approach A hybrid support vector regression (SVR) – seasonal autoregressive integrated moving averages (SARIMA) model is proposed to model the seasonal, linear and nonlinear components of the tourism demand in a destination country. The paper further proposes the use of multiple criteria decision-making (MCDM) approaches in selecting the best forecasting model among a set of considered models. As such, a preference ranking organization method for enrichment of evaluations (PROMETHEE) II is used to rank the considered forecasting models. Findings The proposed hybrid SVR-SARIMA model is the best performing model among a set of considered models in this paper using performance criteria that evaluate the errors of magnitude, directionality and trend change, of a forecasting model. Moreover, the use of the MCDM approach is found to be a relevant and prospective approach in selecting the best forecasting model among a set of models. Originality/value The novelty of this paper lies in several aspects. First, this paper pioneers the demonstration of the SVR-SARIMA model’s capability in forecasting long-term tourism demand. Second, this paper is the first to have proposed and demonstrated the use of an MCDM approach for performing model selection in forecasting. Finally, this paper is one of the very few papers to provide lenses on the current status of Philippine tourism demand.

Estimation of solar radiation using neighboring stations through hybrid support vector regression boosted by Krill Herd algorithm

Solar radiation is a basic input in many fields of studies and models. However, the low density of solar network stations; the improper distribution of these stations; high cost of purchasing, maintaining, and calibrating solar radiation measurement instruments; and frequent errors in the available data are the most important deficiencies in this regard. Thus, researchers are seeking for new and practical methods to estimate solar radiation accurately. The present study aimed to estimate the solar radiation values based on a new hybrid support vector regression model. To this aim, the solar radiation values of all eight target synoptic stations during 1974–2014 were estimated by using Krill-Herd hybrid algorithm (SVR-KHA) method based on support vector regression and implementing neighboring station data. Results indicated that the testing performance of SVR-KHA has a more precision and lower error for all target stations, compared with classical SVR. In addition, the best results were obtained for SVR-KHA3 hybrid model (Isfahan station). Further, the RMSE, MAPE, and R2 values for this model were 1.98 MJ/m2/day, 7.4%, and 0.93, respectively. In accordance with the results, Krill-Herd algorithm method coupled with support vector regression had a high performance and capability for solar radiation estimation in Iran. In other words, the hybrid SVR-KHA model is more flexible and has less error in modeling the nonlinear and complex systems. Finally, the new method of using neighboring stations can be regarded as an appropriate method for estimating nonlinear phenomenon such as solar radiation.

Applying hybrid support vector regression and genetic algorithm to water alternating CO2gas EOR

AbstractWater alternating CO2gas injection (WAG CO2) is one of the most promising enhanced oil recovery techniques. The optimization of this process requires performing many time‐consuming simulations. In this paper, an intelligent hybridization based on support vector regression (SVR) and genetic algorithm (GA) is introduced for the WAG process optimization in the presence of time‐dependent constraints. Multiple SVRs are used as dynamic proxy to mimic numerical simulator behavior in real time. Latin hypercube design (LHD) is applied to generate the proper runs to train the proxy and ten supplementary runs are randomly chosen to validate it. The goal of GA in this study is twofold. First, it is employed during the training of multiple SVRs to find their appropriate hyper‐parameters. Second, once the training and validation of the dynamic proxy are done, the GA is coupled with it to find the optimum WAG parameters which maximize field oil production total (FOPT) subject to time‐dependent water‐cut constraint and some domain constraints. The task is formulated as a non‐linear constrained optimization problem. A semi‐synthetic WAG CO2case is used to examine the reliability of the approach. The results show that the established dynamic proxy is fast and accurate in reproducing the simulator outputs. The hybridization proxy‐GA is demonstrated to be reliable for the real‐time optimization of the formulated WAG process. © 2020 The Authors.Greenhouse Gases: Science and Technologypublished by Society of Chemical Industry and John Wiley & Sons, Ltd.