support vector regression (svr)

Due to the dynamic structure and nonlinear fluctuations of the stock market, it is difficult to accurately predict the trend of this market using the old methods. In this study, in order to improve the accuracy of predicting the index trend in different industries, we propose a new algorithm that combines algorithms fractal interpolation and support vector machine regression, abbreviated as fracsion algorithm. . For this purpose, after recognizing the fractal structure of industries using the Hurst exponent of each industry, we consider the value of the index in each fractal industry as the primary data to predict the trend of the index. Then, by modifying the fractal interpolation algorithm, we will generate new data, and finally, by calling the support vector regression algorithm on the obtained data, we will predict the index trend. The results of the implementation of the Hybrid fracsion algorithm and its comparison with two conventional methods, namely artificial neural network and support vector machine regression, indicate the superiority of the predictive accuracy of the proposed algorithm.

The Karun River is the biggest river basin in Iran, which supplies water demands of about 16 cities, several villages, thousands of hectares of agricultural. This river polluted because of domestic and urban sewerage, industrial sources, and irrigation of agricultural land, Hospital sewage and high tide level of Persian Gulf.
Therefore, because of the importance of this river, the water salinity of this river is determined in this study. The traditional methods of determining water salinity are costly in comparison with remote sensing methods.
In the present study, Landsat 8 (OLI) data was used to calculate the water salinity map for Karun River since not only it is free, but it also has an acceptable resolution.

Landsat 8 (OLI) images were used to calculate reflectance for a pixel and were attained from (US Geological Survey (USGS) 2019). First, radiometric correction was applied to normalize satellite images. This process convert Digital Number into radiance. Second, in order to attain the surface reflectance values, the process of atmospheric correction was applied using the Fast Line-of-sight Atmospheric Analysis of Spectral Hypercubes (FLAASH).
Water salinity was calculate by Iran Water and Power Recourses Development Company. Eight stations are located in the crucial point for EC measuring ALIKALE, GOTVAND, SHOOSHTAR, SHOTEYT, GARGAR, DEZ, AHVAZ, and ABADAN.
Iran Water and Power Recourses Development Company obtained 102 observed EC samples from June 2013 to July 2018 along the Karun River.
The Support Vector Machine was classically used for classification, Support Vector Classification, but extended for using along with regression issue, namely Support Vector Regression.
The results related to the quality of the SVR depend on some factors: the loss function Ɛ, the error penalty factor C and the kernel function parameters.
Usually, radial basis kernel function (RBF), k(x, x΄) = k(x,x΄)=exp⁡〖( -||x-x΄〗 2/σ^2), has been used in remote sensing studies, so, it is implemented in this study. Finally, the Genetic Algorithm (GA) is employed to optimize some parameters including C, Ɛ and σ.
GA is an optimization technique create by Holland (1975) and discussed the mechanism of GA in solving nonlinear optimization problems.
Besides, the Genetic Algorithm (GA) was applied to determine the best performer bands combination. Furthermore, we employ GA to optimize SVR parameters and number of layers and neurons of MLP neural network in order to maximize model accuracy.

Salinity intrusion is a complex issue in coastal, hot, and dry areas. Currently, remote sensing techniques have been widely used to monitor water salinity changes, ranging from inland river networks to deep oceans. The Karun River basin, with a basin area of 67,000 km^2, is located in the southern part of Iran. The salinity of Karun River has been increasing due to some critical factors, e.g. severe climate condition and regional physiography, industrial sources, domestic and urban sewerage, irrigation of agricultural land, fish hatchery, hospital sewage, and high tide level of Persian Gulf .
This study aimed at building Support Vector Regression (SVR) and Multilayer Perceptron (MLP) models to realize the salinity intrusion through the relationship between reflectance from the Landsat-8 Operational Land Imager images and salinity levels measured in situ. 102 observed samples were divided into 75% training and 25% test.
Besides, the Genetic Algorithm (GA) was applied to determine the best performer bands combination. Furthermore, we employ GA to optimize SVR parameters and number of layers and neurons of MLP neural network in order to maximize model accuracy.
The result showed that the MLP approach was the better model to estimate water salinity along the Karun River network in the study area, which coefficient of determination (R^2) and RMSE of test data is obtained as 0.73 and 390μs 〖cm〗^(-1).
GA analysis proved that bands 1, 2 and 3 are the best for modeling water salinity. In this study, the GA is used to determine the SVR meta-parameters including the loss function Ɛ, the error penalty factor C and σ parameters, which are obtained to be〖1×10〗^(-9), 1099 and 0.96, respectively, and number of layers and neurons of MLP neural network, which are obtained to be 5 and 35, respectively.
The result showed that the MLP approach was the better model to estimate water salinity along the Karun River network in the study area, which coefficient of determination (R^2) and RMSE of test data is obtained as 0.73 and 390μs 〖cm〗^(-1).

The present study calculated the relationship between reflectance retrieved from Landsat-8 OLI and water salinity in the Karun River. SVR and MLP models had acceptable operation by considering the large size, geographic complexity of the study domain and the wide range of field data that change between 385 and 4310μs cm^(-1). Augmentation field data is the critical priority work for future study to probe the relationship between water salinity and satellite images.In addition, the contribution of thermal bands can help to increase accuracy of models. Salinity intrusion is a complex issue in coastal and hot and dry areas. Currently, remote sensing techniques have been widely used to monitor water salinity changes, ranging from inland river networks to deep oceans. The Karun River basin, with a basin area of 67,000 km2, is located in the southern part of Iran. The salinity of Karun River has been increasing due to some critical factors, e.g. severe climate condition and regional physiography, industrial sources, domestic and urban sewerage, irrigation of agricultural land, fish hatchery, hospital sewage, and high tide level of Persian Gulf .This study aimed at building Support Vector Regression (SVR) and Multilayer Perceptron (MLP) models to realize the salinity intrusion through the relationship between reflectance from the Landsat-8 Operational Land Imager images and salinity levels measured in situ. A total of 102 observed samples were divided into 75% training and 25% test. Besides, the Genetic Algorithm (GA) was applied to determine the best performer bands combination. Furthermore, we employ GA to optimize SVR parameters and number of layers and neurons of MLP neural network in order to maximize model accuracy. The result showed that the MLP approach was the better model to estimate water salinity along the Karun River network in the study area, which coefficient of determination (R2) and RMSE of test data is obtained as 0.73 and 390μscm-1

Summary: Water saturation (Sw) of a hydrocarbon reservoir is an important petrophysical parameter having a great impact on the accuracy of primitive estimation of the reservoir. Due to highly importance of this parameter dealing with the economic calculations of the reservoir, it must be estimated precisely. Although experimental analysis of core samples taken from a reservoir leads to very useful information about Sw of the reservoir, this experimental method is highly expensive and time consuming; and therefore, this method is applicable only for a small number of wells in a field. To overcome this problem, an intelligent pattern recognition method, known as support vector regression (SVR), has been employed in the current research to estimate Sw from well logs data of 3 wells in one of the largest oil fields of Iran. The performance of the algorithm has also been validated through different criteria. The results of this research indicate that the SVR model can estimate Sw from well logs data accurately, in which the determination coefficients of 87 and 76 percent have been obtained from the training and test steps, respectively. Introduction: Generally in most commonly hydrocarbon reservoirs, Sw is estimated using well logs data through applying Archie's fundamental empirical relation. However, this relation is just satisfied for clean sandstone formations (without clay minerals). So far several empirical models have been proposed to measure Sw using well logs data. The main disadvantage of these models is their formation dependency, which makes the models specific and not comprehensive to be applied in a variety of other formations. In addition to empirical methods, several linear regression techniques have also been applied to estimate this parameter using well logs data. These techniques cannot estimate Sw appropriately due to the complexity of the parameter features. Resistivity and porosity logs are the most important well logs used to estimate Sw by Archie's relation. The porosity of a formation can be very accurately determined through sonic, density and neutron logs. However, resistivity logs are very sensitive to the presence of shale and other clayey impurities in formations. Their effects can be adjusted by means of gamma ray (GR) log. Therefore, to estimate Sw,, employimg an intelligent method using appropriate well logs data will be useful. The oil reservoir, studied in this research, is located in Asmari formation in southwest of Zagros Mountain. Overall this formation in the investigated region has been formed from a sequence comprising of carbonate rocks (limestone and dolomite), sandstone and shale. Methodology and Approaches: In the current research, to estimate Sw, SVR method has been applied to well logs data from 3 wells in one of the largest oil fields of Iran. In this study, appropriate well logs data comprising of GR, neutron porosity, formation bulk density, sonic transit time and true resistivity from deep induction log (ILD) have been used. Moreover, Sw values measured from cores in the laboratory are available for whole depth of the wells. In order to employ SVR to estimate Sw, the model needs to be trained using appropriate input and output data in MATLAB environment. In the current research, the input consists of 5 variables (well logs data) while the output is only the Sw parameter. From 1211 data points (containing 5 variables of well logs data and Sw parameter measured by core) available from the 3 wells, about 80 percent (i.e. 988 samples) were selected for training and the remaining 20 percent (i.e. 223 samples) were chosen for test. To compare the estimated values with the measured ones for the reservoir in the study region, visually, chart of lithology, water and hydrocarbon saturations of the formation were also depicted for the 3 wells by means of Geolog software. Results and Conclusions: The performance of the algorithm has been validated through different criteria such as scatter plot of Sw values from cores versus the estimated Sw values from well logs data of 3 study wells by means of SVR model as well as computing statistical parameters indicating the accuracy of the results. Furthermore, the results of the research revealed that the SVR model can estimate Sw using well logs data accurately so that it has estimated the training and test data with the determination coefficients of 87 and 76 percent, respectively. As a result, the proposed method, i.e. SVR, is an accurate, fast and cost-effective method to evaluate the petrophysical parameter Sw.