Ayşe Berika VAROL MALKOÇOĞLU
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
Ayşe Berika VAROL MALKOÇOĞLU has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Ruya Samli
Istanbul University-Cerrahpasa
2 shared publications - Didem Baran Ergül
Beykoz Üniversitesi
1 shared publication - İsMail İşEri
Ondokuz Mayis University
1 shared publication Ruya Samli
Istanbul University-Cerrahpaşa, Turkey
1 shared publication- Seden Acun Özgünler
Istanbul Technical University
1 shared publication - Zeynep Orman
İstanbul Üniversitesi-Cerrahpaşa
1 shared publication
Latest Publications
- Open Access
- 20 February 2026
- Published by AVES YAYINCILIK A.Ş. in Electrica
- Vol. 26, 1-17
A Detailed Review of YOLO Algorithms: YOLOv1 to YOLOv11
YOLO is a real-time object detection (OD) algorithm with convolution layers that performs OD and classification directly at the image level. Due to the operations performed at the image level, it works very fast compared to its peers. The YOLO algorithm, which is frequently used in both academic and industrial studies thanks to its speed-performance balance, is systematically analyzed from YOLOv1 to YOLOv11 in this study. In the study, the structural components of the algorithms are analyzed and a clear and simple architecture is drawn. Innovations and improvements are mentioned, and the development of the series is explained step by step. Object detection scenarios realized with YOLO in the literature are reviewed and organized into topics. In addition, the performance results in the original papers are compared based on co-metrics and datasets, and the development of the series is comprehensively analyzed. In the last stage, the projected areas of improvement and potential research areas for further development are discussed, and recommendations for researchers are presented. Cite this article as: A. B. Varol Malko.oğlu and R. Şamlı, “A detailed review of YOLO algorithms: YOLOv1 to YOLOv11,” Electrica, 26, 0047, 2026. doi:10.5152/electrica.2026.26047.
- 20 February 2026
- Published by AVES YAYINCILIK A.Ş. in Electrica
- Vol. 26, 1-17
Open AccessA Detailed Review of YOLO Algorithms: YOLOv1 to YOLOv11
YOLO is a real-time object detection (OD) algorithm with convolution layers that performs OD and classification directly at the image level. Due to the operations performed at the image level, it works very fast compared to its peers. The YOLO algorithm, which is frequently used in both academic and industrial studies thanks to its speed-performance balance, is systematically analyzed from YOLOv1 to YOLOv11 in this study. In the study, the structural components of the algorithms are analyzed and a clear and simple architecture is drawn. Innovations and improvements are mentioned, and the development of the series is explained step by step. Object detection scenarios realized with YOLO in the literature are reviewed and organized into topics. In addition, the performance results in the original papers are compared based on co-metrics and datasets, and the development of the series is comprehensively analyzed. In the last stage, the projected areas of improvement and potential research areas for further development are discussed, and recommendations for researchers are presented. Cite this article as: A. B. Varol Malko.oğlu and R. Şamlı, “A detailed review of YOLO algorithms: YOLOv1 to YOLOv11,” Electrica, 26, 0047, 2026. doi:10.5152/electrica.2026.26047.
- Open Access
- 26 September 2024
- Published by Muhendislik Bilimleri ve Tasarim Dergisi in Mühendislik Bilimleri ve Tasarım Dergisi
- Vol. 12 (3) , 488-504
CLASSIFICATION OF ACUTE LYMPHOBLASTIC LEUKEMIA CELLS USING ARTIFICIAL INTELLIGENCE
Due to the morphological similarity between immature lymphoblasts (cancerous cells) to lymphocytes (non-cancerous cells), detecting Acute Lymphoblastic Leukemia poses a significant challenge for pathologists. These cells, which exhibit a similar pattern, can lead to various errors during the diagnosis of the disease. In this study, the cancerous and non-cancerous cells were classified using 3 different artificial intelligence approaches. In the first approach, the classification process was carried out by training Convolutional Neural Networks in 4 different architectures. In the second approach, a hybrid approach was proposed by combining the convolution layer of the CNN model as the feature extractor with the Support Vector Machine, Naive Bayes and Random Forest algorithms as the classifier. The classification processes were carried out by training the proposed second approach. In the third approach, the classification process was performed using transfer learning process and ResNet50 and VGG16 networks. In all experiments, the effects of hyper-parameter and dataset changes on model performance were also examined. The results obtained by these three approaches were compared using the Accuracy, Precision, Recall, F-score, and AUC performance measures. It was determined that the most successful results were obtained with the 1st approach using the Dataset3.
- 26 September 2024
- Published by Muhendislik Bilimleri ve Tasarim Dergisi in Mühendislik Bilimleri ve Tasarım Dergisi
- Vol. 12 (3) , 488-504
Open AccessCLASSIFICATION OF ACUTE LYMPHOBLASTIC LEUKEMIA CELLS USING ARTIFICIAL INTELLIGENCE
Due to the morphological similarity between immature lymphoblasts (cancerous cells) to lymphocytes (non-cancerous cells), detecting Acute Lymphoblastic Leukemia poses a significant challenge for pathologists. These cells, which exhibit a similar pattern, can lead to various errors during the diagnosis of the disease. In this study, the cancerous and non-cancerous cells were classified using 3 different artificial intelligence approaches. In the first approach, the classification process was carried out by training Convolutional Neural Networks in 4 different architectures. In the second approach, a hybrid approach was proposed by combining the convolution layer of the CNN model as the feature extractor with the Support Vector Machine, Naive Bayes and Random Forest algorithms as the classifier. The classification processes were carried out by training the proposed second approach. In the third approach, the classification process was performed using transfer learning process and ResNet50 and VGG16 networks. In all experiments, the effects of hyper-parameter and dataset changes on model performance were also examined. The results obtained by these three approaches were compared using the Accuracy, Precision, Recall, F-score, and AUC performance measures. It was determined that the most successful results were obtained with the 1st approach using the Dataset3.
- 30 July 2023
- Published by Mimarlik Bilimleri ve Uygulamalari Dergisi in Mimarlık Bilimleri ve Uygulamaları Dergisi (MBUD)
- Vol. 8 (1) , 15-24
Yapılarda Yeni İşlev Seçimi İçin Bulanık Mantık Tabanlı Karar Destek Sistemlerinin Kullanımı
These buildings, which function as important documents for the period in which they were built, and which have managed to preserve their original qualities, have been abandoned as a result of various some factors and have lose their functionality. It is very important to continue to use these original structures, which have since lost their functionality, with re-functioning as opposed to a passive conservation approach. Architectural decision support systems, used to create a solution mechanism can assist architects in deciding on the most appropriate function option for building by systematizing the relationship between spatial analysis of the existing building stock and functional expectations. Accordingly, the results were observed by using fuzzy logic method in order to determine the most suitable function option before the design and construction phases for reuse of the building were started.
- 30 July 2023
- Published by Mimarlik Bilimleri ve Uygulamalari Dergisi in Mimarlık Bilimleri ve Uygulamaları Dergisi (MBUD)
- Vol. 8 (1) , 15-24
Yapılarda Yeni İşlev Seçimi İçin Bulanık Mantık Tabanlı Karar Destek Sistemlerinin Kullanımı
These buildings, which function as important documents for the period in which they were built, and which have managed to preserve their original qualities, have been abandoned as a result of various some factors and have lose their functionality. It is very important to continue to use these original structures, which have since lost their functionality, with re-functioning as opposed to a passive conservation approach. Architectural decision support systems, used to create a solution mechanism can assist architects in deciding on the most appropriate function option for building by systematizing the relationship between spatial analysis of the existing building stock and functional expectations. Accordingly, the results were observed by using fuzzy logic method in order to determine the most suitable function option before the design and construction phases for reuse of the building were started.
- Open Access
- 29 June 2023
- Published by Turkiye Bilisim Vakfi in Türkiye Bilişim Vakfı Bilgisayar Bilimleri ve Mühendisliği Dergisi
- Vol. 16 (1) , 60-80
Video Verilerinde Bulunan Tehlikeli Nesnelerin Derin Öğrenme Yöntemleri ile Tespiti Üzerine Derleme
Bilgisayarla görme tekniklerinden biri olan nesne saptaması son yıllarda hem akademik hem de ticarî potansiyeli sayesinde büyük ilgi görmektedir. Günümüzde teknolojinin gelişimi ile birlikte güvenlik ya da kişisel amaçlarla çekilen video görüntülerinin artması ve donanım elemanlarının gelişmesi, ihtiyaç duyulan kaynaklara erişimi kolaylaştırmış dolayısıyla nesne saptama sistemlerinin gelişimini hızlandırmıştır. Bu alanda yaya saptaması, yüz tanıma gibi bazı klasikleşmiş konularda çok sayıda çalışma bulunmaktadır. Fakat bu çalışmada farklı nesne gruplarının getirdiği zorlukları gözlemlemek adına tehlikeli nesneler üzerine yapılan ve güvenlik güçlerine yardımcı sistemlerin tasarlanmasına katkı sağlayan çalışmalar araştırılıp derlenmiştir. Çalışmalarda kullanılan nesne saptama yöntemleri geleneksel yöntemler ve derin öğrenme tabanlı modern yöntemler olarak iki kısımda incelenmiş olup avantajları ve dezavantajları tartışılmıştır. Ayrıca literatürdeki eksiklikler belirlenip, gelecekteki çalışmalar için araştırmacılara yönergeler sunulmuştur.
- 29 June 2023
- Published by Turkiye Bilisim Vakfi in Türkiye Bilişim Vakfı Bilgisayar Bilimleri ve Mühendisliği Dergisi
- Vol. 16 (1) , 60-80
Open AccessVideo Verilerinde Bulunan Tehlikeli Nesnelerin Derin Öğrenme Yöntemleri ile Tespiti Üzerine Derleme
Bilgisayarla görme tekniklerinden biri olan nesne saptaması son yıllarda hem akademik hem de ticarî potansiyeli sayesinde büyük ilgi görmektedir. Günümüzde teknolojinin gelişimi ile birlikte güvenlik ya da kişisel amaçlarla çekilen video görüntülerinin artması ve donanım elemanlarının gelişmesi, ihtiyaç duyulan kaynaklara erişimi kolaylaştırmış dolayısıyla nesne saptama sistemlerinin gelişimini hızlandırmıştır. Bu alanda yaya saptaması, yüz tanıma gibi bazı klasikleşmiş konularda çok sayıda çalışma bulunmaktadır. Fakat bu çalışmada farklı nesne gruplarının getirdiği zorlukları gözlemlemek adına tehlikeli nesneler üzerine yapılan ve güvenlik güçlerine yardımcı sistemlerin tasarlanmasına katkı sağlayan çalışmalar araştırılıp derlenmiştir. Çalışmalarda kullanılan nesne saptama yöntemleri geleneksel yöntemler ve derin öğrenme tabanlı modern yöntemler olarak iki kısımda incelenmiş olup avantajları ve dezavantajları tartışılmıştır. Ayrıca literatürdeki eksiklikler belirlenip, gelecekteki çalışmalar için araştırmacılara yönergeler sunulmuştur.
- 27 May 2022
- Published by IGI Global Scientific Publishing
- p. 156-171
Earthquake Risk Prediction With Artificial Intelligence Methods
Earthquakes are one of the most difficult natural phenomena in human history to predict. Today, despite very advanced technologies, earthquake predictions still have not been conclusive. It is especially known that the trilogy of location, time, and magnitude is quite difficult to predict at the same time. In order to discover this powerful natural phenomenon, scientists are trying to collect and make sense of the parameters affecting the earthquake and the earthquake results. In general, their goal is to determine the characteristics that have an impact on earthquakes, to perform classifications thanks to various artificial intelligence algorithms, and to predict future earthquakes. The aim of this study is to compile, examine, and analyze earthquake risk prediction researches or applications carried out using artificial intelligence methods. The studies obtained as a result of the literature review were grouped according to the metrics used, data sets, features, and models used and evaluated according to the success rates obtained.
- 27 May 2022
- Published by IGI Global Scientific Publishing
- p. 156-171
Earthquake Risk Prediction With Artificial Intelligence Methods
Earthquakes are one of the most difficult natural phenomena in human history to predict. Today, despite very advanced technologies, earthquake predictions still have not been conclusive. It is especially known that the trilogy of location, time, and magnitude is quite difficult to predict at the same time. In order to discover this powerful natural phenomenon, scientists are trying to collect and make sense of the parameters affecting the earthquake and the earthquake results. In general, their goal is to determine the characteristics that have an impact on earthquakes, to perform classifications thanks to various artificial intelligence algorithms, and to predict future earthquakes. The aim of this study is to compile, examine, and analyze earthquake risk prediction researches or applications carried out using artificial intelligence methods. The studies obtained as a result of the literature review were grouped according to the metrics used, data sets, features, and models used and evaluated according to the success rates obtained.