Since the formation of the Iranian electricity market in 1382 (2003), power plants have been competing with each other on a daily basis in the ISO by registering their bid prices. In this competition, the winners are those power plants whose bid prices are lower than the market clearing price for each hour in the next day, so the forecasting the next day market prices is vital for energy producers. In this study, using a combination of K-means algorithm and support vector machine, a new model for predicting the next day market settlement prices is proposed and the model has been used the hourly electricity market prices for 1395-1396 (2016-2017). According to the results, seven competitive clusters were identified for the Iranian electricity market. The average forecasting accuracy of the proposed model for each of these clusters for the years 1395 (2016) and 1396 (2017) was 96% and 94%, repectively.

The purpose of this paper is to predict stock prices using Hybrid GA-SVM algorithm. Predicting time series such as stock price forecasting is one of the most important issues in financial field. In real life, identifying time series movements in stock price indices is very complex. Therefore, the use of a classical model alone cannot accurately predict stock price indices. Hence, by using combined methods, uncertainty in forecasting can be reduced. In stock price forecasting in financial sector, more than 100 indicators have been created to understand stock market behavior, so, identifying the appropriate indicators is a challenging problem. One of the techniques that has recently been studied for serial forecasting is support regression Vector (SVR) or machine support vector (SVM). This study uses the GA-SVM hybrid algorithm to predict the stock price index. Experimental results show that Hybrid GA-SVM algorithm provides a more appropriate and promising alternative to stock market forecasting.

Recent significant development in sensor technology makes possible Earth observational remote sensing system with unprecedented spectral resolution and data dimensionality. The value of these new sensor systems lies in their ability to acquire a nearly complete optical spectrum for each pixel in the scene. Such imaging spectrometry now makes possible the acquisition of data in hundreds of spectral bands simultaneously, and it is called hyperspectral images. With the limited number of training samples of hyperspectral images, the classification of these images using conventional feature extraction algorithms (PCA, ICA, PP, DBFE, DAFE and Wavelet) is considered useless. In this paper a two stages classification algorithm is proposed, by fussing the spatial and spectral information. In the first stage the classes of each pixel and its eight neighbors are identified, using a classical classification algorithm. In the second stage two primary classes of a pixel and its neighbors are compared in each node of decision tree by a SVM. The proposed, binary tree SVM, takes advantage of both the efficient computation of the tree architecture and the high classification accuracy of SVM. The hyperspectral data set used in our experiments is a scene from Indiana’s Indian Pine by the AVIRIS sensor. The examples results show the problem of limited training samples can be mitigated using the proposed algorithm; moreover the computational time is significantly reduced. This suggests that binary tree SVM could be a promising tool for classifying hyperspectral images.

Classification of high-dimensional hyperspectral data with many spectral bands for the derivation of good accuracy is an important problem in hyperspectral remote sensing. The most of classification algorithms are based on spectral information. Here, in order to achieve an high classification accuracy, we can use the spatial information of data. Integration of hidden morkov random field that optimize spatial information by minimizing energy functions, with support vector machine that is an powerful method for classification of hyperspectral data, can improve classification accuracy in final classified map properly. The purpose of this study is to improve the classification accuracy with a limited of training samples by combination of support vector machine algorithm and hidden morkov random field. In this study, tow hyperspectral dataset from Hyperion and AVIRIS sensors has been used. After the applying radiometric corrections like correcting embedded lines and remove bad bands, atmospheric correction Hyperion dataset done by FLAASH method and AVIRIS dataset by IAR algorithm. MNF transformation was used in order to dimensionally reduction and the endmembers were extracted from PPI band and then in order to spectral classification, used from SVM method. Finally, to improve classification accuracy in the final classified map, hidden Markov random field (HMRF) was used. So that after the extracting of Components from PCA and MNF Transformations, computing of some statistic parameters of classes in SVM classified map in order to use in inputs model and so configuration of iterations, SVM-HMRF model was applied. The results show that the proposed model (SVM-HMRF) has improved overall classification accuracy in both of data sets. For example, the improved classification accuracy on some of land uses, were around 25 percent. Also regions of final classified map is much more homogeneous and salt and pepper nose drastically reduced.

Background: A brain tumor is one of the most common and fatal neurological diseases that may require surgery. The correct diagnosis of the location and size of the tumor can be a diagnostic aid program for medical robots during surgery, and it also helps doctors formulate a suitable treatment plan for the patient. Objectives: To develop an algorithm based on support vector machine (SVM) for the detection and classification of tumors into benign and malignant types on MRI images. Methods: In this retrospective study, 160 MRI images were obtained from the KAGGLE website. The studied subjects included two groups: Benign tumors and malignant tumors. At first, preprocessing and noise removal were done by comparing four filters: Butterworth, wavelet, ideal, and median. Finally, the SVM algorithm was used to classify brain tumors into benign and malignant. Results: The performance evaluation of the filters showed that the median and wavelet filters had the best performance in removing noise from MRI images. Then, the discrete wavelet transform (DWT) extracted the required features from MRI images and was used as the input of the SVM algorithm. The accuracy, precision and specificity of the proposed algorithm in diagnosing benign and malignant brain tumors were 95%, 88% and 91%. Conclusions: The findings of recent studies show that this algorithm can be used to improve the accurate diagnosis of brain tumors and their types. Combining morphological features can also be a diagnostic tool to increase accuracy in robotic surgeries.

Computer networks are spreading widely and one of the most outstanding challenges in computer network security is detecting intrusions into networks. One of the main tools for detection is controlling network traffic and analyzing users’ behavior. One way of accomplishing this is to set classifications that specify the patterns in huge volumes of data. By means of data mining methods and introducing a binary label (normal pack, abnormal pack) and specifying the priority of data, abnormal data is detected leading to increased accuracy of network intrusion detection which in turn leads to improvement and maintenance of network security. In this paper, SVM algorithm is analyzed in terms of priorities and the effect of machine learning algorithm on accuracy of intrusion detection is investigated. The results show that using SVM is more advantageous compared to past approaches yielding better detection and increasing accuracy and right alarm detection.

We present a novel algorithm, which is called Cutting algorithm (CA), for improving the accuracy and reducing the computations of the Least Squares Support Vector Machines (LS-SVMs). The method is based on dividing the original problem to some subproblems. Since a master problem is converted to some small problems, so this algorithm has fewer computations. Although, in some cases that the typical LS-SVM cannot classify the dataset linearly, applying the CA the datasets can be classified. In fact, the CA improves the accuracy and reduces the computations. The reported and comparative results on some known datasets and synthetics data demonstrate the efficiency and the performance of CA.

In this research, the performance of support vector machine in predicting relative energy dissipation in non-prismatic channel and rough bed with trapezoidal elements has been investigated. To achieve the objectives of the present study, 136 series of laboratory data are analyzed under the same laboratory conditions using a support vector machine. The present study entered the support vector machine network without dimension in two different scenarios with a height of 1. 50 and 3. 0 cm rough elements. Two statistical criteria of Root Mean Square Error and coefficient of determination are used to evaluate the efficiency of input compounds. Hydraulically, the results show that at both heights of the rough elements, energy dissipation increased with increasing Froude number. The results of the support vector machine show that the height of the roughness element is 1. 50 cm in the first scenario, combination number 6 with R2 = 0. 990 and RMSE = 0. 0129 for training mode and R2 = 0. 993 and RMSE = 0. 032 for testing mode and the height of the roughness element 3. 0 in the second scenario, combination number 6 with R2 = 0. 989 and RMSE = 0. 0112 for training mode, R2 = 0. 994 and RMSE = 0. 0224 for testing mode are select as the best models. Finally, sensitivity analysis is performed on the parameters and H / y1 parameter is selected as the most effective parameter.