Transactions on Machine Intelligence2821-16931120180301Spam Detection from Big Data based on Evolutionary Data Mining Systems1915971510.47176/TMI.2018.1ENH.Ehsani ChimehDepartment of Electrical Engineering, Amirkabir University of Technology, Tehran, IranM.KaramiDepartment of Electrical Engineering, Shahid Beheshti University, Tehran, IranJournal Article20180129News releases and users' ability to discuss events, events, and writing personalities and environments are services that provide opportunities for new types of spam and spammers. For example, popular topics and topics that involve the most discussions can be an opportunity to create traffic, visits, and sources of income. When something happens, thousands of users write about it, send text and quickly become the subject of discussion. These topics are targeted by spammers, because their writings contain the common words used in popular discussions. Often there are links in spam that direct users to websites that are not related to the topic, and since these URLs are shortened, it's difficult for users to log in. This type of spams can reduce the value and efficiency of instantaneous search services, and users of these services refer to materials that do not contain links to the searcher, so a method for identifying spammers should be found. Methods available to deal with spammers can be included in three categories which contain detection-based approach, prevention-based approach, and degradation-based approach that this research uses is a detection approach. Hence, this research uses a smart method that initially enters large data into the program, then a feature extraction based on the genetic algorithm is performed. In the next step, the classification of data in order to detect spam is done using the combined method of self-organized mapping neural network and probabilistic neural network with the support vector machine core as a radial basis function.https://www.tmachineintelligence.ir/article_159715_eee24d1d83b39451758f002bb11cb928.pdfTransactions on Machine Intelligence2821-16931120180301Sensorless Stator Flux Oriented Control for Startup Gas Turbine101815971610.47176/TMI.2018.10ENF.MohammadiSchool of Electrical and Computer Engineering University of TehranF. R.SalmasiSenior Member, IEEE School of Electrical and Computer Engineering University of TehranJournal Article20180120This paper introduces a novel hybrid rotor‐position estimation technique for wound‐field synchronous motors (WFSM) powered by load‐commutated inverters (LCI), operating continuously from standstill through turbine ignition speed. At low speeds, where signal‐injection techniques excel, a high‐frequency injection‐based estimator provides initial rotor‐angle information. As speed increases past this low‐speed region, a model‐based stator‐flux observer seamlessly takes over both position estimation and firing‐pulse generation for the thyristor bridge. By combining these two complementary estimation strategies, the proposed algorithm overcomes limitations inherent in each individual method—namely, poor low‐speed observability for flux observers and the energy losses associated with continuous injection at higher speeds. The result is reliable commutation of the load‐commutated inverter’s thyristors across the entire speed range, from zero to nominal. In addition, we present a new stator‐flux‐oriented control architecture tailored for LCI‐fed WFSMs. This structure enhances the motor’s power factor and yields a faster dynamic response in the speed controller, while simplifying the overall drive design to reduce manufacturing costs. The efficacy of the hybrid estimation algorithm and the novel control structure is demonstrated through detailed MATLAB/Simulink simulations. Results confirm seamless transition between estimation modes, robust commutation under varying operating conditions, improved power quality, and accelerated speed regulation—validating the approach as a practical, cost‐effective solution for high‐performance WFSM drives.https://www.tmachineintelligence.ir/article_159716_cab293851798f958be178f21aff8b1a3.pdfTransactions on Machine Intelligence2821-16931120180301The Best Planning For a Grid-Connected Microgrid Takes Into Account Load and Renewable Generation Uncertainty As Well As Battery Storage193015971810.47176/TMI.2018.19ENA.RezazadehDepartment of electrical and computer engineering, Shahid Beheshti University, Tehran, IranH. R.BagheriDepartment of electrical and computer engineering, Shahid Beheshti University, Tehran, IranA.A.SarabadaniDepartment of electrical and computer engineering, Shahid Beheshti University, Tehran, IranJournal Article20180116One of the finest solutions for supplying electrical energy in rural places is to use hybrid renewable energy. When using renewable energy sources to meet demand, the right capacity of these sources should be chosen because they are dependent on weather and other factors. It is very impressive to take into account the stochastic nature of wind speed and solar radiation when estimating the potential of renewable energy sources like wind and solar. One issue with employing renewable energy like wind and solar in micro-grids is their inherent unpredictability and random stochastic nature, which made planning and forecasting for such resources challenging. To represent uncertainty in both Wind and PV resources, stochastic programming and probability scenarios are used in this project. Gams software uses mixed integer programming to determine the best way to program a micro-grid that is connected to the main grid. The Virtual Power Producer uses the main control system to manage optimal production and load control.https://www.tmachineintelligence.ir/article_159718_d389f5781d9040c60fd839507be047dc.pdfTransactions on Machine Intelligence2821-16931120180301Economic Analysis of Off-Grid Reverse Osmosis Desalination: A Case Study in Chabahar, Iran313815972010.47176/TMI.2018.31ENS.Rasi ErshadiDepartment of Electrical Engineering, Iran University of Science And Technology, Tehran, IranR.MiremadiDepartment of Mechanical Engineering, Shahid Beheshti University, Tehran, IranH.ShayanfarDepartment of Electrical Engineering, Iran University of Science And Technology, Tehran, IranJournal Article20220109Accessing fresh water in Chabahar, Iran’s only oceanic port, poses significant challenges due to both geographic and infrastructural constraints. Situated along the Gulf of Oman, Chabahar faces acute water scarcity, making the deployment of a water desalination system an attractive and necessary solution to meet the city’s growing water demands. Among available technologies, reverse osmosis (RO) desalination stands out as an efficient and widely adopted method for seawater purification. However, the successful implementation of such a system requires careful consideration of its power supply, especially given the lack of a reliable local power infrastructure. This paper presents a comprehensive techno-economic analysis of an RO water desalination system tailored for Chabahar’s unique conditions. To address the energy needs, various distributed generation (DG) options are evaluated to ensure continuous and cost-effective electrical supply to the desalination units. System design and optimal component sizing are conducted using the HOMER software, which allows for detailed modeling under different operational scenarios. The technical findings are then integrated into a financial analysis using COMFAR software, enabling assessment of investment feasibility, operating costs, and long-term sustainability. Notably, real-world regional data are incorporated throughout the analysis to enhance the study’s practical relevance and applicability, offering valuable insights for policymakers and planners.https://www.tmachineintelligence.ir/article_159720_57a1a599540f1ebfc9c834cb657c3e2e.pdfTransactions on Machine Intelligence2821-16931120180301On-Line Reusing-Based Scheduling Algorithm for 2-Dimensional Tasks in Reconfigurable Hardware394815972310.47176/TMI.2018.39ENA.ParisayDepartment of Electrical Engineering, University of Science and Technology (IUST), Tehran, IranH.Shahriar ShahhoseiniDepartment of Electrical Engineering, University of Science and Technology (IUST), Tehran, IranS. M.MohtavipourDepartment of Electrical Engineering, University of Science and Technology (IUST), Tehran, IranJournal Article20180119Reducing reconfiguration overhead is critical to improving the runtime performance of dynamically reconfigurable Field-Programmable Gate Arrays (FPGAs). In this paper, we introduce a novel task-reuse strategy tailored for two-dimensional FPGA hardware layouts. The key idea is to identify and exploit repetitive computations by reusing already‐configured hardware modules rather than incurring costly bitstream reloads. First, incoming tasks are classified into two categories significant (high‐impact or frequently appearing) and non-significant based on metrics such as execution frequency, resource intensity, and temporal locality. Each category is assigned to its own hardware partition. Within the significant partition, when a new significant task arrives, the system either replaces an existing module whose future utility is low or, if sufficient empty regions exist in the non-significant partition, temporarily maps the task there. If neither option is feasible, the partition boundary is extended to accommodate the new module, up to predefined physical limits. We evaluated our approach on a suite of benchmark applications exhibiting high task repetition. Compared to leading dynamic‐reconfiguration algorithms, our method reduced overall makespan by 20.3% on average. Moreover, the task-placement decision algorithm operates in polynomial time, achieving placement decisions over three times faster than competing strategies. These results demonstrate that intelligent partitioning combined with selective reuse and partition‐border extension can substantially lower reconfiguration overhead and accelerate FPGA‐based computation pipelines.https://www.tmachineintelligence.ir/article_159723_9e40291e8def2a62d9ccfc7db4c5ad58.pdfTransactions on Machine Intelligence2821-16931120180301Intelligent Transportation in The Prevention of Accidents495615982010.47176/TMI.2018.49ENS.AsadzadehDepartment of Computer Engineering, Marvdasht Branch, Islamic Azad University, Marvdasht, IranO.Rahmani SeryasatAssistant Prof, Department of Electrical Engineering, Shams Higher Education Institute, Gorgan, Iran0000-0002-9289-6128S.MoshrefzadehDepartment of Computer Engineering, Faculty of Technology and Engineering, Yasouj University, Yasouj, IranJournal Article20220116Information technology has significantly influenced numerous industrial sectors, and its integration into transportation systems has emerged as a promising solution, giving rise to intelligent transportation systems (ITS). Among the key application areas of ITS is the use of computer vision for accident prevention. Specifically, intelligent driver monitoring systems play a crucial role in enhancing vehicle safety by proactively identifying and addressing conditions that may lead to accidents. These systems aim to assist and alert drivers by intelligently recognizing potentially dangerous situations, thereby contributing to a substantial reduction in traffic accidents and related incidents. A major concern within intelligent transportation is driver drowsiness, a critical factor in preventing severe financial and human losses caused by traffic accidents. To address this, intelligent systems are employed to enhance vehicle control by continuously analyzing the driver’s physical state and behavioral patterns. Consequently, the development of a system capable of accurately assessing a driver’s alertness or fatigue level based on both driver behavior and vehicle status holds great importance. Notably, the proposed system presented in this research demonstrates superior performance compared to existing approaches, achieving 96% accuracy, 94% sensitivity, and 94% specificity in detecting driver drowsiness.https://www.tmachineintelligence.ir/article_159820_839da70edc02b98667e62acbd9da358d.pdf