Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collections

Permanent URI for this collectionhttps://hdl.handle.net/20.500.13091/3

Browse

Now showing 1 - 20 of 206

Accurate Edge Detection With Support of Reflectance Transformation Imaging
(Institute of Electrical and Electronics Engineers Inc., 2022) Kaya, Burhan; Durdu, Akif
Reflectance Transformation Imaging (RTI) is a method of photographing an object that reveals details that are invisible to the naked eye. The input of RTI consists of a series of images captured by a fixed positioned camera and each illuminated from a known and different direction by lights. Reflection Transform Imaging is widely used to produce quality models from multi-light image data. It is frequently preferred for various studies in the field of cultural heritage. For the first time in this paper, the RTI photographing method has been used outside of its traditional using way. It is used to solve the well-known problem of edge detection. Reflection transform cannot be used actively, because it is difficult to create an RTI experimental environment in daily life. However, under certain conditions, the approaches mentioned in this paper will be used in daily life, from the analysis of images in every field. In this paper, the ideas that it can be applied in every partially controlled area that needs high resolution object detection are discussed. With the method mentioned in this statement, a new approach has been proposed and proven based on RTI basics for edge and corner detections. © 2022 IEEE.
Citation - WoS: 3
Citation - Scopus: 4
Achievable Rate Analysis for Two-Way Relay Non-Orthogonal Multiple Access Systems
(IEEE, 2021) Özdemir, Özgür
This paper investigates the performance of a non-orthogonal multiple access (NOMA) based two-way relaying system where the users want to exchange independent messages with the help of a decode-and-forward relay. We consider transmission over three phases where the first and second phases are allocated to the transmissions of the users and after detection the relay applies superposition coding and transmits the network encoded symbol to the users in the third phase. Exact analytical expressions are derived to characterize the achievable average rate of the system over independent Rayleigh fading channels. Computer simulations are also presented to confirm the theoretical analysis. Analytical and simulation results show that the proposed three-phase two-way relaying scheme with NOMA outperforms the two-phase and four-phase NOMA-based two-way relaying scenarios in terms of achievable average rate.
Citation - Scopus: 1
Achievable Rate of Noma-Based Cooperative Communication Systems With Best Relay Selection Over Cascaded Rayleigh Fading Channels
(IEEE, 2020) Özdemir, Özgür
In this paper, the achievable rate analysis of NOMA-based cooperative communication systems with best relay selection is studied. The cascaded Rayleigh fading channels are considered since investigations have shown that cascaded channel structure agree better with mobile network models such as inter-vehicular communication systems. A cooperative network where a source terminal communicates with a destination directly and through a selected relay among K relays is considered and the achievable average rate of this system using NOMA is found by computer simulations. The obtained results for cascaded Rayleigh fading channels in case of decode and forward protocol have shown that the average rate is decreased as the cascading degree increases. It has been also seen that for a given cascading degree the average rate performance of the system is increased when the number of total relays is increased.
Citation - WoS: 24
Citation - Scopus: 40
An Adaptive Method for Traffic Signal Control Based on Fuzzy Logic With Webster and Modified Webster Formula Using Sumo Traffic Simulator
(IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2021) Ali, Muzamil Eltejani Mohammed; Durdu, Akif; Çeltek, Seyit Alperen; Yılmaz, Alper
In the past, the Webster optimal cycle time formula was limited to calculate the optimal cycle from historical data for fixed-time traffic signal control. This paper focuses on the design of an adaptive traffic signal control based on fuzzy logic with Webster and modified Webster's formula. These formulas are used to calculate the optimal cycle time depending on the current traffic situation which applying in the next cycle. The alternation of the traffic condition between two successive cycles is monitored and handled through the fuzzy logic system to compensate the fluctuation. The obtained optimal cycle time is used to determine adaptively the effective phase green times i.e. is used to determine adaptively the maximum allowable extension limit of the green phase in the next cycle. The SUMO traffic simulator is used to compare the results of the proposed adaptive control methods with fuzzy logic-based traffic control, and fixed-time Webster and modified Webster-based traffic control methods. The proposed methods are tested on an isolated intersection. In this study, real field-collected data obtained from three, four, and five approaches intersections in Kilis/Turkey are used to test the performance of the proposed methods. In addition, to examine the efficiency of the proposed techniques at heavy demands, the arbitrary demands are generated by SUMO for a four approaches intersection. The obtained simulation results indicate that the proposed methods overperform the fixed time and fuzzy logic-based traffic control methods in terms of average vehicular delay, speed, and travel time.
Citation - WoS: 1
Citation - Scopus: 2
Adaptive State Feedback Control Method Based on Recursive Least Squares
(Kauno Technologijos Universitetas, 2022) Levent, Mehmet Latif; Aydoğdu, Ömer
This study revealed an adaptive state feedback control method based on recursive least squares (RLS) that is introduced for a time-varying system to work with high efficiency. Firstly, a system identification block was created that gives the mathematical model of the time-varying system using the input/output data packets of the controller system. Thanks to this block, the system is constantly monitored to update the parameters of the system, which change over time. Linear quadratic regulator (LQR) is renewed according to these updated parameters, and self-adjustment of the system is provided according to the changed system parameters. The Matlab/Simulink state-space model of the variable loaded servo (VLS) system module was obtained for the simulation experiments in this study; then the system was controlled. Moreover, experiments were carried out on the servo control experimental equipment of the virtual simulation laboratories (VSIMLABS). The effectiveness of the proposed new method was observed taking the performance indexes as a reference to obtain the results of the practical application of the proposed method. Regarding the analysis of simulation and experimental results, the proposed approach minimizes the load effect and noise and the system works at high efficiency. © 2022 Kauno Technologijos Universitetas. All rights reserved.
Citation - WoS: 5
Citation - Scopus: 6
Adrenal Tumor Characterization on Magnetic Resonance Images
(WILEY, 2020) Barstuğan, Mücahid; Ceylan, Rahime; Asoğlu, Semih; Cebeci, Hakan; Koplay, Mustafa
Adrenal tumors occur on adrenal glands and are generally detected on abdominal area scans. Adrenal tumors, which are incidentally detected, release vital hormones. These types of tumors that can be malignant affect body metabolism. Both of benign and malign adrenal tumors can have a similar size, intensity, and shape, this situation may lead to wrong decision during diagnosis and characterization of tumors. Thus, biopsy is done to confirm diagnosis of tumor types. In this study, adrenal tumor characterization is handled by using magnetic resonance images. In this way, it is wanted that patient can be disentangled from one or more imaging modalities (some of them can includes X-ray) and biopsy. An adrenal tumor image set, which includes five types of adrenal tumors and has 112 benign tumors and 10 malign tumors, was used in this study. Two data sets were created from the adrenal tumor image set by manually/semiautomatically segmented adrenal tumors and feature sets of these data sets are constituted by different methods. Two-dimensional gray-level co-occurrence matrix (2D-GLCM), gray-level run-length matrix (GLRLM), and two-dimensional discrete wavelet transform (2D-DWT) methods were analyzed to reveal the most effective features on adrenal tumor characterization. Feature sets were classified in two ways: benign/malign (binary classification) and type characterization (multiclass classification). Support vector machine and artificial neural network classified feature sets. The best performance on benign/malign classification was obtained by the 2D-GLCM feature set. The best results were assessed with sensitivity, specificity, accuracy, precision, and F-score metrics and they were 99.17%, 90%, 98.4%, 99.17%, and 99.13%, respectively. The highest classification performance on type characterization was obtained by the 2D-DWT feature set as 59.62%, 96.17%, 93.19%, 54.69%, and 54.94% for sensitivity, specificity, accuracy, precision, and F-score metrics, respectively.
Adrenal Tumor Classification on T1 and T2-Weighted Abdominal Mr Images
(Institute of Electrical and Electronics Engineers Inc., 2019) Barstuğan, Mücahid; Ceylan, Rahime; Asoğlu, Semih; Cebeci, Hakan; Koplay, Mustafa
Adrenal tumors occur on adrenal glands and can be malignant. Adrenal glands consist of cortex and medulla. If cortex or medulla produce hormones extremely, the hormonal unbalance situation arises. This situation causes adrenal tumor occurrence on adrenal glands. In this study, adrenal tumors on T1 and T2-weighted MR images were classified by the SVM algorithm. Before the classification stage, different feature extraction algorithms and filtering methods were used for preprocessing. The classification results that were obtained by four different methods were evaluated on five different evaluation metrics as sensitivity, specificity, accuracy, precision, and F-score. The best classification performance was obtained with Method 2 on T1-weighted MR (Magnetic Resonance) images where the sensitivity, specificity, accuracy, precision, and F-score metrics were obtained as 99.17%, 90%, 98.4%, 99.17%, and 99.13%, respectively. © 2019 IEEE.
Citation - WoS: 9
Citation - Scopus: 11
Adrenal Tumor Segmentation Method for Mr Images
(ELSEVIER IRELAND LTD, 2018) Barstuğan, Mücahid; Ceylan, Rahime; Asoğlu, Semih; Cebeci, Hakan; Koplay, Mustafa
Background and objective: Adrenal tumors, which occur on adrenal glands, are incidentally determined. The liver, spleen, spinal cord, and kidney surround the adrenal glands. Therefore, tumors on the adrenal glands can be adherent to other organs. This is a problem in adrenal tumor segmentation. In addition, low contrast, non-standardized shape and size, homogeneity, and heterogeneity of the tumors are considered as problems in segmentation. Methods: This study proposes a computer-aided diagnosis (CAD) system to segment adrenal tumors by eliminating the above problems. The proposed hybrid method incorporates many image processing methods, which include active contour, adaptive thresholding, contrast limited adaptive histogram equalization (CLAHE), image erosion, and region growing. Results: The performance of the proposed method was assessed on 113 Magnetic Resonance (MR) images using seven metrics: sensitivity, specificity, accuracy, precision, Dice Coefficient, Jaccard Rate, and structural similarity index (SSIM). The proposed method eliminates some of the discussed problems with success rates of 74.84%, 99.99%, 99.84%, 93.49%, 82.09%, 71.24%, 99.48% for the metrics, respectively. Conclusions: This study presents a new method for adrenal tumor segmentation, and avoids some of the problems preventing accurate segmentation, especially for cyst-based tumors. (C) 2018 Elsevier B.V. All rights reserved.
Citation - WoS: 8
Citation - Scopus: 10
Analysis and Design of a Transimpedance Amplifier Based Front-End Circuit for Capacitance Measurements
(SPRINGER INTERNATIONAL PUBLISHING AG, 2020) Demirtaş, Mehmet; Erişmiş, Mehmet Akif; Güneş, Salih
In this study, transimpedance amplifier based front-end circuits which can be employed to measure small capacitances were designed, analyzed and simulated using analog electronic circuit simulator. The front-end circuit converts the current flowing through the measured capacitance into a modulated voltage value which contains information regarding the desired capacitance. The frequency-domain, time-domain, stability and noise analyzes were carried out numerically and in simulation environment using a circuit simulator. The analytical, numerical and simulation results can be used to design optimized, precise and stable transimpedance amplifiers with low-noise value. The measured capacitance value was 10 pF which is low enough to simulate various real-world applications. Three commercially available, off-the-shelf operational amplifiers with different peripheral passive components were employed for computer based analysis. The designed transimpedance amplifiers are suitable to connect with capacitance extraction circuits which use analog or digital demodulation techniques.
Citation - Scopus: 20
Analysis and Investigation of Different Advanced Control Strategies for High-Performance Induction Motor Drives
(Universitas Ahmad Dahlan, 2020) Farah, Nabil; Talib, Hairul Nizam; Ibrahim, Z.B.; Abdullah, Q.; Aydoğdu, Ömer; Rasin, Z.; Lazi, J.M.
Induction motor (IM) drives have received a strong interest from researchers and industry particularly for high-performance AC drives through vector control method. With the advancement in power electronics and digital signal processing (DSP), high capability processors allow the implementation of advanced control techniques for motor drives such as model predictive control (MPC). In this paper, design, analysis and investigation of two different MPC techniques applied to IM drives; the model predictive torque control (MPTC) and model predictive current control (MPCC) are presented. The two techniques are designed in Matlab/Simulink environment and compared in term of operation in different operating conditions. Moreover, a comparison of these techniques with field-oriented control (FOC) and direct torque control (DTC) is conducted based on simulation studies with PI speed controller for all control techniques. Based on the analysis, the MPC techniques demonstrates a better result compared with the FOC and DTC in terms of speed, torque and current responses in transient and steady-state conditions. © (2020). All rights reserved.
Citation - Scopus: 7
Analysis of Average Bit Error Rate for Ofdm-Im Systems With Hardware Impairments Over Nakagami-M and Weibull Fading Channels
(Institute of Electrical and Electronics Engineers Inc., 2022) Ceniklioğlu, Büşra; Develi, I.; Canbilen, Ayşe Elif; Lafcı, Mustafa
Orthogonal frequency division multiplexing together with index modulation (OFDM-IM) is a brilliant key alternative to the traditional orthogonal frequency division multiplexing (OFDM) schemes for future communication technologies whereby it has high spectral and energy efficiency. However, there are still key issues that need to be adressed to evaluate possible applications in practical systems. Accordingly, we investigate the performance of OFDM-IM systems over Nakagami-m and Weibull fading channels that suffer from transceiver hardware impairments (HWIs) in this paper. Specifically, a maximum likelihood (ML) detector is applied for OFDM-IM-based transmission and the system performance is analyzed under the effect of HWIs considering average bit error rate (ABER) by utilizing computer simulations. The obtained results show that the HWIs have serious destructive effects on the overall system performance. © 2022 IEEE.
Citation - WoS: 13
Citation - Scopus: 18
Application of the Maturity Model in Industrial Corporations
(Mdpi, 2022) Ünal, Cihan; Sungur, Cemil; Yıldırım, Hakan
Enterprises need to evaluate for themselves whether they are ready for Industry 4.0 to survive and develop in the era of the Fourth Industrial Revolution. Therefore, it is necessary to conceptualize or develop an Industry 4.0 readiness and maturity model with basic model dimensions. The present study aimed to review the maturity models available in the literature and to develop and implement a comprehensive maturity model that would eliminate the problems in the existing models. Most maturity models developed lack vital dimensions such as laws, incentives, and corporate culture. While developing the model, AHP and expert opinions were used to determine the dimension weights. The model was applied to 87 businesses in various industries at the Ankara Chamber of Industry Industrial Park in Turkey. The developed model calculates the maturity level of the enterprise for six dimensions. The data on 61 corporations where Industry 4.0 technologies were adopted were analyzed based on demographic variables such as the year of establishment, industry, size, capital, and turnover. These findings demonstrated that Industry 4.0 was introduced recently in Turkey and businesses are required to take further steps to keep up with the global digital transformation. Since the number of industries and corporations that are aware of the Industry 4.0 technologies is limited in Ankara, Turkey, only a few businesses adopted the Industry 4.0 technologies. This developed model will make an important contribution to the literature with its unique dimensions. It would pave the way for further research in various industries in Turkey and other nations where Industry 4.0 investments are new.
Citation - WoS: 1
Citation - Scopus: 3
An Approach for Learning From Robots Using Formal Languages and Automata
(EMERALD GROUP PUBLISHING LTD, 2019) Aslan, Muhammet Fatih; Durdu, Akif; Sabancı, Kadir; Erdogan, Kemal
Purpose In this study, human activity with finite and specific ranking is modeled with finite state machine, and an application for human-robot interaction was realized. A robot arm was designed that makes specific movements. The purpose of this paper is to create a language associated to a complex task, which was then used to teach individuals by the robot that knows the language. Design/methodology/approach Although the complex task is known by the robot, it is not known by the human. When the application is started, the robot continuously checks the specific task performed by the human. To carry out the control, the human hand is tracked. For this, the image processing techniques and the particle filter (PF) based on the Bayesian tracking method are used. To determine the complex task performed by the human, the task is divided into a series of sub-tasks. To identify the sequence of the sub-tasks, a push-down automata that uses a context-free grammar language structure is developed. Depending on the correctness of the sequence of the sub-tasks performed by humans, the robot produces different outputs. Findings This application was carried out for 15 individuals. In total, 11 out of the 15 individuals completed the complex task correctly by following the different outputs. Originality/value This type of study is suitable for applications to improve human intelligence and to enable people to learn quickly. Also, the risky tasks of a person working in a production or assembly line can be controlled with such applications by the robots.
Citation - WoS: 1
Citation - Scopus: 2
Artificial Intelligence Based High Voltage Cable Bonding To Prevent Cable Termination Faults
(ELSEVIER SCIENCE SA, 2020) Akbal, Bahadır
Cable termination fault (CTF) is a major problem for high voltage cable lines (HVCL). Increasing of the sheath voltage (SV), zero sequence current (ZC) and current harmonic distortion (THDI) on metallic sheath (MS) of HVC are major factors for CTF. MS is grounded according to IEEE 575-1988 standard to reduce SV. However, these methods are not sufficient to prevent CTF based on ZC and THDI. The aims of this paper are minimization of SV, ZC and THDI to prevent CTF based on ZC and THDI. Thus, LSSB method is developed as a new bonding method. Also, LSSB parameters should be optimized to make the most economical and practical bonding. GA, DEA, PSO and iPSO are used optimization methods for optimization of LSSB. SV and THDI should be known for optimization of LSSB, so the forecasting methods (FM) are used as fitness function of optimization methods in LSSB optimization. The regression and hybrid artificial neural network methods are compared to determine the most suitable FM. When the optimized LSSB is used for bonding of long HVCL, SV reduces approximately 90%, ZC reduces approximately 93%, and THDI reduces approximately 70%. Thus CTF risk is minimized by using the optimized LSSB in HVCL.
Citation - WoS: 6
Citation - Scopus: 7
Automated Elimination of Eog Artifacts in Sleep Eeg Using Regression Method
(2019) Dursun, Mehmet; Özşen, Seral; Güneş, Salih; Akdemir, Bayram; Yosunkaya, Şebnem
Sleep electroencephalogram (EEG) signal is an important clinical tool for automatic sleep staging process. Sleep EEG signal is effected by artifacts and other biological signal sources, such as electrooculogram (EOG) and electromyogram (EMG), and since it is effected, its clinical utility reduces. Therefore, eliminating EOG artifacts from sleep EEG signal is a major challenge for automatic sleep staging. We have studied the effects of EOG signals on sleep EEG and tried to remove them from the EEG signals by using regression method. The EEG and EOG recordings of seven subjects were obtained from the Sleep Research Laboratory of Meram Medicine Faculty of Necmettin Erbakan University. A dataset consisting of 58 h and 6941 epochs was used in the research. Then, in order to see the consequences of this process, we classified pure sleep EEG and artifact-eliminated EEG signals with artificial neural networks (ANN). The results showed that elimination of EOG artifacts raised the classification accuracy on each subject at a range of 1%– 1.5%. However, this increase was obtained for a single parameter. This can be regarded as an important improvement if the whole system is considered. However, different artifact elimination strategies combined with different classification methods for another sleep EEG artifact may give higher accuracy differences between original and purified signals.
Citation - WoS: 25
Citation - Scopus: 36
Automatic Detection of Power Quality Disturbance Using Convolutional Neural Network Structure With Gated Recurrent Unit
(Hindawi Limited, 2021) Yiğit, E.; Özkaya, U.; Öztürk, Ş.; Singh, D.; Gritli, H.
Power quality disturbance (PQD) is essential for devices consuming electricity and meeting today's energy trends. This study contains an effective artificial intelligence (AI) framework for analyzing single or composite defects in power quality. A convolutional neural network (CNN) architecture, which has an output powered by a gated recurrent unit (GRU), is designed for this purpose. The proposed framework first obtains a matrix using a short-time Fourier transform (STFT) of PQD signals. This matrix contains the representation of the signal in the time and frequency domains, suitable for CNN input. Features are automatically extracted from these matrices using the proposed CNN architecture without preprocessing. These features are classified using the GRU. The performance of the proposed framework is tested using a dataset containing a total of seven single and composite defects. The amount of noise in these examples varies between 20 and 50 dB. The performance of the proposed method is higher than current state-of-the-art methods. The proposed method obtained 98.44% ACC, 98.45% SEN, 99.74% SPE, 98.45% PRE, 98.45% F1-score, 98.19% MCC, and 93.64% kappa metric. A novel power quality disturbance (PQD) system has been proposed, and its application has been represented in our study. The proposed system could be used in the industry and factory. © 2021 Enes Yi?it et al.
Automatic Phase Reversal Detection in Routine Eeg
(CHURCHILL LIVINGSTONE, 2020) Yıldırım, Sema; Koçer, Hasan Erdinç; Ekmekçi, Ahmet Hakan
Electroencephalograph (EEG), a valuable tool in the clinical evaluation, is readily available, safe and provides information about brain function. EEG interpretation is important for the diagnosis of neurological disorders. The long-term EEG data may be required to document and study neurosciences that include many epileptic activities and phase reversal (PR) etc. However, analyze of the long-term EEG done by an expert neurologist is much time consuming and quite difficult. Therefore, an automatic PR determination method for analyzing of long-term EEG is described in this study. The presented technique was applied to the pathological EEG recordings that were obtained from two different datasets gathered as a retrospective in Selcuk University Hospital (SUH) and Boston Children's Hospital (BCH). With this method, PR in the dataset was determined and then compared with the ones detected by the specialist doctor. Two tests were carried out in the SUH dataset and the classification success of the method was 83.22% for test 1 and 85.19% for test 2. On the other hand, three tests were carried out for two different position values for BCH dataset. The highest classification success of the six tests was 75% for test 5, while the lowest classification success appeared as 58.33% for test 6. As a result, the overall success in the detection of PR with the conducted method is 84.20% for SUH and 66.7% for BCH. According to these results, the determination of PR that is known to be indicative of neurological disorders and presenting them to expert information will accelerate the interpretation of long-term EEG recordings.
Citation - WoS: 2
Citation - Scopus: 2
Autonomously Simultaneous Localization and Mapping Based on Line Tracking in a Factory-Like Environment
(VSB-TECHNICAL UNIV OSTRAVA, 2019) Durdu, Akif; Korkmaz, Mehmet
This study is related to SLAM, also known simultaneous localization and mapping which is highly important and an indispensable issue for autonomous mobile robots. Both an environment mapping and an agent's localization are provided with SLAM systems. However, while performing SLAM for an unknown environment, the robot is navigated by three different ways: a user guidance, random movements on an exploration mode or exploration algorithms. A user guidance or random exploration methods have some drawbacks that a user may not be able to observe the agent or random process may take a long time. In order to answer these problems, it is searched for a new and autonomous exploration algorithm for SLAM systems. In this manner, a new kind of left-orientated autonomous exploration algorithm for SLAM systems has been improved. To show the algorithm effectiveness, a factorylike environment is made up on the ROS (Robot Operating System) platform and navigation of the agent is observed. The result of the study demonstrates that it is possible to perform SLAM autonomously in any similar environment without the need of the user interference.
Citation - WoS: 19
Citation - Scopus: 24
Bacteria Foraging Optimisation Algorithm Based Optimal Control for Doubly-Fed Induction Generator Wind Energy System
(INST ENGINEERING TECHNOLOGY-IET, 2020) Bakır, Hale; Merabet, Adel; Dhar, Rupak Kanti; Kulaksız, Ahmet Afşin
In this study, an optimisation method, based on bacteria foraging, is investigated to tune the parameters of the proportional-integral (PI) controllers in a doubly-fed induction generator (DFIG) wind energy system connected to the grid. The generator is connected to the grid directly at the stator and through the back-to-back converter at the rotor. The control system includes PI controllers, at the rotor side, to regulate the rotor currents and PI controller to regulate the dc-link voltage for efficient power transfer. The control parameters, of three PI controllers, are optimised offline using the bacteria foraging optimisation algorithm and a modelled DFIG wind energy system. Various performance criteria, based on the tracking errors, are used to assess the efficiency of the optimisation method. Furthermore, the conventional tuning method and genetic algorithm optimisation method are conducted and compared to the bacteria foraging optimisation method to demonstrate its advantages. The optimised control parameters are evaluated on a DFIG wind energy experimental setup. Experimental and simulation results are provided to validate the effectiveness of each optimisation method.
Citation - WoS: 21
Citation - Scopus: 24
Boosting the Oversampling Methods Based on Differential Evolution Strategies for Imbalanced Learning
(Elsevier, 2021) Korkmaz, Sedat; Sahman, Mehmet Akif; Çınar, Ahmet Cevahir; Kaya, Ersin
The class imbalance problem is a challenging problem in the data mining area. To overcome the low classification performance related to imbalanced datasets, sampling strategies are used for balancing the datasets. Oversampling is a technique that increases the minority class samples in various proportions. In this work, these 16 different DE strategies are used for oversampling the imbalanced datasets for better classification. The main aim of this work is to determine the best strategy in terms of Area Under the receiver operating characteristic (ROC) Curve (AUC) and Geometric Mean (G-Mean) metrics. 44 imbalanced datasets are used in experiments. Support Vector Machines (SVM), k-Nearest Neighbor (kNN), and Decision Tree (DT) are used as a classifier in the experiments. The best results are produced by 6th Debohid Strategy (DSt6), 1th Debohid Strategy (DSt1), and 3th Debohid Strategy (DSt3) by using kNN, DT, and SVM classifiers, respectively. The obtained results outperform the 9 state-of-the-art oversampling methods in terms of AUC and G-Mean metrics (C) 2021 Elsevier B.V. All rights reserved.

Browse

Browsing Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collections by Department "Fakülteler, Mühendislik ve Doğa Bilimleri Fakültesi, Elektrik-Elektronik Mühendisliği Bölümü"