Gürkan Doğan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Gürkan Doğan has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Burhan Ergen
Firat Universitesi
11 shared publications - Abdulkadir Sengur
Department of Electrical-Electronic Engineering, Faculty of Technology, Firat University, 23000 Elazig, Turkey
2 shared publications - Andaç Imak
Department of Electrical and Electronic Engineering, Faculty of Engineering, Munzur University, 62000 Tunceli, Turkey
2 shared publications - Aybars Mahmat
Munzur Üniversitesi
2 shared publications - Hakan Uyanik
Electrical-Electronics Engineering Department, Engineering Faculty, Munzur University, Tunceli, Trkiye
2 shared publications - Nihat Tosun
Firat Universitesi
2 shared publications
Latest Publications
- 30 September 2026
- Published by Firat Universitesi in Fırat Üniversitesi Mühendislik Bilimleri Dergisi
- Vol. 38 (2) , 765-782
Ti6AI7Nb Alaşımının Mikro Frezelenmesinde Kesme Kuvvetinin Öngörülmesi için Yenilikçi Hibrit Derin Sinir Ağı Modeli
Titanyum alaşımları, yüksek mukavemet-ağırlık oranları, korozyon dirençleri ve biyouyumlulukları sayesinde havacılıktan savunmaya, biyomedikalden uzay teknolojilerine kadar pek çok kritik alanda vazgeçilmez malzemeler haline gelmiştir. Ancak bu alaşımlar, düşük ısıl iletkenlik ve yüksek kimyasal reaktivite gibi özellikleri nedeniyle işlenmesi güç malzemeler olarak bilinmektedir. Bu durum, özellikle, kesme kuvveti, takım ömrünü, boyutsal hassasiyeti ve iş parçası bütünlüğünü doğrudan etkileyen temel bir parametre olmasından dolayı kesme kuvvetinin doğru tahmin edilmesini hem teknik açıdan zorlaştırmakta hem de süreç güvenilirliği açısından zorunlu kılmaktadır. Bu çalışmada, Ti6AI7Nb alaşımının mikro frezelenmesi sürecinde kesme kuvvetinin yüksek doğrulukla tahmin edilmesine yönelik yeni bir hibrit derin sinir ağı modeli önerilmektedir. Önerilen modelde, evrişimli sinir ağı katmanı ile veriden anlamlı yerel özellikler otomatik olarak çıkarılırken, bu katmana paralel olarak konumlandırılan tam bağlantılı katmanlarda karmaşık ilişkileri ve küresel bilgileri çıkarmak üzere entegre edilmiştir. Önerilen model ve güncel yöntemlerin eğitimi ve testi için kısıtlı olan gerçek deneysel veriler, veri artırma tekniğiyle çoğaltılarak sentetik veriler üretilmiştir. Gerçekleştirilen kapsamlı analizlerde, önerilen paralel hibrit derin öğrenme modelinin kesme kuvveti tahmininde %97.13 R² skoru ile geleneksel makine öğrenmesi yöntemlerinden daha yüksek performans sergilediği ortaya koyulmuştur.
- 30 September 2026
- Published by Firat Universitesi in Fırat Üniversitesi Mühendislik Bilimleri Dergisi
- Vol. 38 (2) , 765-782
Ti6AI7Nb Alaşımının Mikro Frezelenmesinde Kesme Kuvvetinin Öngörülmesi için Yenilikçi Hibrit Derin Sinir Ağı Modeli
Titanyum alaşımları, yüksek mukavemet-ağırlık oranları, korozyon dirençleri ve biyouyumlulukları sayesinde havacılıktan savunmaya, biyomedikalden uzay teknolojilerine kadar pek çok kritik alanda vazgeçilmez malzemeler haline gelmiştir. Ancak bu alaşımlar, düşük ısıl iletkenlik ve yüksek kimyasal reaktivite gibi özellikleri nedeniyle işlenmesi güç malzemeler olarak bilinmektedir. Bu durum, özellikle, kesme kuvveti, takım ömrünü, boyutsal hassasiyeti ve iş parçası bütünlüğünü doğrudan etkileyen temel bir parametre olmasından dolayı kesme kuvvetinin doğru tahmin edilmesini hem teknik açıdan zorlaştırmakta hem de süreç güvenilirliği açısından zorunlu kılmaktadır. Bu çalışmada, Ti6AI7Nb alaşımının mikro frezelenmesi sürecinde kesme kuvvetinin yüksek doğrulukla tahmin edilmesine yönelik yeni bir hibrit derin sinir ağı modeli önerilmektedir. Önerilen modelde, evrişimli sinir ağı katmanı ile veriden anlamlı yerel özellikler otomatik olarak çıkarılırken, bu katmana paralel olarak konumlandırılan tam bağlantılı katmanlarda karmaşık ilişkileri ve küresel bilgileri çıkarmak üzere entegre edilmiştir. Önerilen model ve güncel yöntemlerin eğitimi ve testi için kısıtlı olan gerçek deneysel veriler, veri artırma tekniğiyle çoğaltılarak sentetik veriler üretilmiştir. Gerçekleştirilen kapsamlı analizlerde, önerilen paralel hibrit derin öğrenme modelinin kesme kuvveti tahmininde %97.13 R² skoru ile geleneksel makine öğrenmesi yöntemlerinden daha yüksek performans sergilediği ortaya koyulmuştur.
- 15 September 2026
- Published by Black Sea Journal of Engineering and Science in Black Sea Journal of Engineering and Science
- Vol. 9 (5) , 2652-2662
A New Deep Convolutional Neural Network Model for Multi-Class Cyber Attack Detection in Internet of Medical Things Networks
The Internet of Medical Things (IoMT) has become an indispensable component of modern healthcare, with a wide range of applications from patient monitoring systems to smart medical devices. However, the limited processing power and memory capacity of IoMT devices make it difficult to implement traditional security mechanisms, leaving these devices vulnerable to cyber threats such as protocol attacks (e.g., DDoS, DoS, MQTT) and identity spoofing. Deep learning-based attack detection systems are gaining increasing importance in developing effective solutions against these attacks, which can directly threaten patient safety. In this study, DCNN—a new deep convolutional neural network-based model—is proposed to develop an effective attack detection system against the growing cyberattack threats in IoMT network environments. The model was evaluated on the CICIoMT2024 dataset, which consists of six classes: Benign, DDoS, DoS, MQTT, Reconnaissance, and Spoofing. The performance of the proposed DCNN was compared with re-implemented traditional machine learning models, including Random Forest, LightGBM, Decision Tree, Extra Trees, Logistic Regression, Naive Bayes, and Ridge Classifier, as well as state-of-the-art methods reported in the literature. The experimental results demonstrate that the proposed DCNN outperformed all comparison models, achieving 0.9999 or higher across all evaluation metrics, including accuracy, F1-score, precision, recall, and AUC.
- 15 September 2026
- Published by Black Sea Journal of Engineering and Science in Black Sea Journal of Engineering and Science
- Vol. 9 (5) , 2652-2662
A New Deep Convolutional Neural Network Model for Multi-Class Cyber Attack Detection in Internet of Medical Things Networks
The Internet of Medical Things (IoMT) has become an indispensable component of modern healthcare, with a wide range of applications from patient monitoring systems to smart medical devices. However, the limited processing power and memory capacity of IoMT devices make it difficult to implement traditional security mechanisms, leaving these devices vulnerable to cyber threats such as protocol attacks (e.g., DDoS, DoS, MQTT) and identity spoofing. Deep learning-based attack detection systems are gaining increasing importance in developing effective solutions against these attacks, which can directly threaten patient safety. In this study, DCNN—a new deep convolutional neural network-based model—is proposed to develop an effective attack detection system against the growing cyberattack threats in IoMT network environments. The model was evaluated on the CICIoMT2024 dataset, which consists of six classes: Benign, DDoS, DoS, MQTT, Reconnaissance, and Spoofing. The performance of the proposed DCNN was compared with re-implemented traditional machine learning models, including Random Forest, LightGBM, Decision Tree, Extra Trees, Logistic Regression, Naive Bayes, and Ridge Classifier, as well as state-of-the-art methods reported in the literature. The experimental results demonstrate that the proposed DCNN outperformed all comparison models, achieving 0.9999 or higher across all evaluation metrics, including accuracy, F1-score, precision, recall, and AUC.
- 4 September 2026
- Published by Elsevier in Internet of Things
- Vol. 40
FedMedIDS: A non-IID-resilient federated DCNN framework for securing IoMT and IoHT devices
- 4 September 2026
- Published by Elsevier in Internet of Things
- Vol. 40
FedMedIDS: A non-IID-resilient federated DCNN framework for securing IoMT and IoHT devices
- 1 July 2026
- Published by Elsevier in Journal of Power Sources
- Vol. 690
Modeling stack voltage and remaining useful life of proton exchange membrane fuel cells using deep neural networks
- 1 July 2026
- Published by Elsevier in Journal of Power Sources
- Vol. 690
Modeling stack voltage and remaining useful life of proton exchange membrane fuel cells using deep neural networks
- 2 January 2026
- Published by International Journal of Assessment Tools in Education in International Journal of Assessment Tools in Education
- Vol. 13 (1) , 186-204
Predicting middle school students’ academic achievement in mathematics with a new hybrid approach
The academic achievement of middle school students is one of the critical steps in their educational journey that affects their future academic and career prospects. During this transition period, students are faced with a variety of challenges related to their families, schools, and the individual themselves. Research shows that academic performance during these formative years is critical for future educational success and overall life success. Therefore, addressing students’ academic achievement, which depends on various factors during this period, is of particular importance for both institutional and individual planning and orientation in the future. In this study, a new hybrid approach based on artificial neural networks that enables automatic analysis of data on family, school, and individual factors affecting middle school students’ academic achievement in mathematics is proposed. A publicly available student performance dataset was used for training and testing the proposed hybrid approach and other models. This dataset consists of data such as mathematics grades, family information, residential information, and health status information for 395 students enrolled in two public schools in the Alentejo region of Portugal. The proposed approach achieved an R2 score of 88.6% in experimental studies with this data set, providing approximately 3% higher accuracy than its closest competitor among other methods in the literature.
- 2 January 2026
- Published by International Journal of Assessment Tools in Education in International Journal of Assessment Tools in Education
- Vol. 13 (1) , 186-204
Predicting middle school students’ academic achievement in mathematics with a new hybrid approach
The academic achievement of middle school students is one of the critical steps in their educational journey that affects their future academic and career prospects. During this transition period, students are faced with a variety of challenges related to their families, schools, and the individual themselves. Research shows that academic performance during these formative years is critical for future educational success and overall life success. Therefore, addressing students’ academic achievement, which depends on various factors during this period, is of particular importance for both institutional and individual planning and orientation in the future. In this study, a new hybrid approach based on artificial neural networks that enables automatic analysis of data on family, school, and individual factors affecting middle school students’ academic achievement in mathematics is proposed. A publicly available student performance dataset was used for training and testing the proposed hybrid approach and other models. This dataset consists of data such as mathematics grades, family information, residential information, and health status information for 395 students enrolled in two public schools in the Alentejo region of Portugal. The proposed approach achieved an R2 score of 88.6% in experimental studies with this data set, providing approximately 3% higher accuracy than its closest competitor among other methods in the literature.
- Open Access
- 29 August 2025
- Published by Academician Publishing
AI-Powered Segmentation of Retinal Fundus Images: Optic Disc and Optic Cup Focus
Accurate segmentation of optic disc (OD) and optic cup (OC) from retinal fundus images plays a crucial role in the early diagnosis and treatment of eye diseases that can lead to irreversible eye damages and blindness, such as glaucoma and diabetic retinopathy, and other diseases. In addition to image processing methods such as thresholding, edge detection, region-based methods for segmentation of these structures, machine learning (ML) based methods have been added in recent years. With deep learning (DL)-based methods, these segmentation processes are much more successful and faster than conventional methods. Also they can be performed independently of clinical conditions. In this book we investigate deep learning methods for segmentation OD and OC structures from retinal fundus images. In addition, in the later parts of the book, a new deep learning model inspired by the U-Net architecture, one of the most popular models in image segmentation, is presented. The proposed model includes modified channel attention, modified spatial attention mechanisms and modified ConvMixer architectures. While it is aimed to perform a more precise segmentation process with the help of the attention mechanism methods, it is aimed to perform these operations both faster and with higher accuracy with the help of the ConvMixer structure. To validate our proposal, we used the DDRD Net dataset, which was created from four different popular fundus datasets. The dataset is publicly available and was created by labeling 1310 colored and equally sized retinal fundus images by experts. The evaluated experiment results show that the proposed model outperforms 8 state-of-the-art models (UNet, SegNet, HRNet, DeeplabV3+, ENet, PSPNet, ICNet, FCN8) which gives the most successful results in image segmentation in terms of all obtained metrics. Obtained results show that the proposed model can be used effectively in the field of ophthalmology to segment the OD and OC on the retinal fundus images.
- 29 August 2025
- Published by Academician Publishing
Open AccessAI-Powered Segmentation of Retinal Fundus Images: Optic Disc and Optic Cup Focus
Accurate segmentation of optic disc (OD) and optic cup (OC) from retinal fundus images plays a crucial role in the early diagnosis and treatment of eye diseases that can lead to irreversible eye damages and blindness, such as glaucoma and diabetic retinopathy, and other diseases. In addition to image processing methods such as thresholding, edge detection, region-based methods for segmentation of these structures, machine learning (ML) based methods have been added in recent years. With deep learning (DL)-based methods, these segmentation processes are much more successful and faster than conventional methods. Also they can be performed independently of clinical conditions. In this book we investigate deep learning methods for segmentation OD and OC structures from retinal fundus images. In addition, in the later parts of the book, a new deep learning model inspired by the U-Net architecture, one of the most popular models in image segmentation, is presented. The proposed model includes modified channel attention, modified spatial attention mechanisms and modified ConvMixer architectures. While it is aimed to perform a more precise segmentation process with the help of the attention mechanism methods, it is aimed to perform these operations both faster and with higher accuracy with the help of the ConvMixer structure. To validate our proposal, we used the DDRD Net dataset, which was created from four different popular fundus datasets. The dataset is publicly available and was created by labeling 1310 colored and equally sized retinal fundus images by experts. The evaluated experiment results show that the proposed model outperforms 8 state-of-the-art models (UNet, SegNet, HRNet, DeeplabV3+, ENet, PSPNet, ICNet, FCN8) which gives the most successful results in image segmentation in terms of all obtained metrics. Obtained results show that the proposed model can be used effectively in the field of ophthalmology to segment the OD and OC on the retinal fundus images.
- 28 July 2025
- Published by Springer Nature in Health Information Science and Systems
- Vol. 13 (1) , 1-20
HASK-Net: new hybrid attention selective kernel network for automatic colon cancer detection from colonoscopy images
Colorectal cancer is a frequently analyzed cancer type seen today. It is the third leading of cancer in terms of incidence. It is among the second leading cases resulting in death. However, when detecting pre-cancerous polyps using traditional methods, the rate of being overlooked is high. Therefore, the rate of missed polyps can be greatly reduced with deep learning-based computer-aided diagnosis (CADx) methods. In this study, a novel convolutional neural network based on CADx, called HASK-Net, is proposed for automatic polyp detection. While HASK-Net uses ResNet50 as a feature extractor in the encoder network, the novely designed hybrid attention selective kernel convolution uses it in the decoder network to increase the representation power in feature maps and learn more complex features. In experimental results on the publicly available Hiper-Kvasir and EndoTech 2020 benchmark datasets, HASK-Net showed extremely promising performance with a dice similarity score of 97.02% and 84.83% and an mIoU score of 95.38% and 84.32%, respectively.
- 28 July 2025
- Published by Springer Nature in Health Information Science and Systems
- Vol. 13 (1) , 1-20
HASK-Net: new hybrid attention selective kernel network for automatic colon cancer detection from colonoscopy images
Colorectal cancer is a frequently analyzed cancer type seen today. It is the third leading of cancer in terms of incidence. It is among the second leading cases resulting in death. However, when detecting pre-cancerous polyps using traditional methods, the rate of being overlooked is high. Therefore, the rate of missed polyps can be greatly reduced with deep learning-based computer-aided diagnosis (CADx) methods. In this study, a novel convolutional neural network based on CADx, called HASK-Net, is proposed for automatic polyp detection. While HASK-Net uses ResNet50 as a feature extractor in the encoder network, the novely designed hybrid attention selective kernel convolution uses it in the decoder network to increase the representation power in feature maps and learn more complex features. In experimental results on the publicly available Hiper-Kvasir and EndoTech 2020 benchmark datasets, HASK-Net showed extremely promising performance with a dice similarity score of 97.02% and 84.83% and an mIoU score of 95.38% and 84.32%, respectively.
- 14 May 2025
- Published by Elsevier in Information Sciences
- Vol. 717
A new lightweight convolutional neural network model for detecting drivable road regions
- 14 May 2025
- Published by Elsevier in Information Sciences
- Vol. 717
A new lightweight convolutional neural network model for detecting drivable road regions
- 13 March 2025
- Published by SAGE Publications in Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science
- Vol. 239 (12) , 4581-4592
Modeling and analysis of WEDM with a new deep neural network
In some cases, the models targeted in machining processes are difficult to apply to industrial production. Artificial intelligence techniques applied to industrial production methods in recent times eliminate these problems in machining. The use of deep learning methods in solving these problems has recently increased. In this study, a new deep neural network (DNN) based on artificial neural networks is proposed to model the machining of AISI4140 steel by wire electrical discharge machining (WEDM). The proposed DNN model predicts surface roughness (SR) and wire wear ratio (WWR), which are the most important inferences in cutting with WEDM. Obtaining inferences such as SR and WWR in a real experimental environment means a lot of time and cost. With the proposed model, there is a significant reduction in time and cost factors. The dataset used to train and test the proposed model was obtained from the real experimental environment. For training and testing the model, the data set is divided into two parts: 20% test data and 80% training data. In the experimental tests carried out with the proposed model, error rates such as Mean Absolute Error (MAE), R 2 , and Mean Squared Error (MSE) accuracy rate between SR and WWR values estimated according to the SR and WWR values obtained from the real environment were examined. MAE, MSE, and R 2 for SR reached 0.0495, 0.0005, and 98.48%, respectively, while MAE, MSE, and R 2 for WWR reached 0.0156, 0.0036, and 98.39%, respectively. These results of the proposed model provided a lower error rate and higher accuracy than the state-of-the-art models in the literature.
- 13 March 2025
- Published by SAGE Publications in Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science
- Vol. 239 (12) , 4581-4592
Modeling and analysis of WEDM with a new deep neural network
In some cases, the models targeted in machining processes are difficult to apply to industrial production. Artificial intelligence techniques applied to industrial production methods in recent times eliminate these problems in machining. The use of deep learning methods in solving these problems has recently increased. In this study, a new deep neural network (DNN) based on artificial neural networks is proposed to model the machining of AISI4140 steel by wire electrical discharge machining (WEDM). The proposed DNN model predicts surface roughness (SR) and wire wear ratio (WWR), which are the most important inferences in cutting with WEDM. Obtaining inferences such as SR and WWR in a real experimental environment means a lot of time and cost. With the proposed model, there is a significant reduction in time and cost factors. The dataset used to train and test the proposed model was obtained from the real experimental environment. For training and testing the model, the data set is divided into two parts: 20% test data and 80% training data. In the experimental tests carried out with the proposed model, error rates such as Mean Absolute Error (MAE), R 2 , and Mean Squared Error (MSE) accuracy rate between SR and WWR values estimated according to the SR and WWR values obtained from the real environment were examined. MAE, MSE, and R 2 for SR reached 0.0495, 0.0005, and 98.48%, respectively, while MAE, MSE, and R 2 for WWR reached 0.0156, 0.0036, and 98.39%, respectively. These results of the proposed model provided a lower error rate and higher accuracy than the state-of-the-art models in the literature.
- Open Access
- 23 February 2024
- Published by Springer Nature in International Journal of Multimedia Information Retrieval
- Vol. 13 (1) , 1-11
A new CNN-based semantic object segmentation for autonomous vehicles in urban traffic scenes
Semantic segmentation is the most important stage of making sense of the visual traffic scene for autonomous driving. In recent years, convolutional neural networks (CNN)-based methods for semantic segmentation of urban traffic scenes are among the trending studies. However, the methods developed in the studies carried out so far are insufficient in terms of accuracy performance criteria. In this study, a new CNN-based semantic segmentation method with higher accuracy performance is proposed. A new module, the Attentional Atrous Feature Pooling (AAFP) Module, has been developed for the proposed method. This module is located between the encoder and decoder in the general network structure and aims to obtain multi-scale information and add attentional features to large and small objects. As a result of experimental tests with the CamVid data set, an accuracy value of approximately 2% higher was achieved with a mIoU value of 70.59% compared to other state-of-art methods. Therefore, the proposed method can semantically segment objects in the urban traffic scene better than other methods.
- 23 February 2024
- Published by Springer Nature in International Journal of Multimedia Information Retrieval
- Vol. 13 (1) , 1-11
Open AccessA new CNN-based semantic object segmentation for autonomous vehicles in urban traffic scenes
Semantic segmentation is the most important stage of making sense of the visual traffic scene for autonomous driving. In recent years, convolutional neural networks (CNN)-based methods for semantic segmentation of urban traffic scenes are among the trending studies. However, the methods developed in the studies carried out so far are insufficient in terms of accuracy performance criteria. In this study, a new CNN-based semantic segmentation method with higher accuracy performance is proposed. A new module, the Attentional Atrous Feature Pooling (AAFP) Module, has been developed for the proposed method. This module is located between the encoder and decoder in the general network structure and aims to obtain multi-scale information and add attentional features to large and small objects. As a result of experimental tests with the CamVid data set, an accuracy value of approximately 2% higher was achieved with a mIoU value of 70.59% compared to other state-of-art methods. Therefore, the proposed method can semantically segment objects in the urban traffic scene better than other methods.