Gülnur Yildizdan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Gülnur Yildizdan has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
- Emine Baş
Konya Teknik Üniversitesi
7 shared publications Emine Baş
1 shared publication- Ömer Kaan Baykan
Computer Engineering Department, Konya Technical University, Konya, Turkey
1 shared publication
Latest Publications
- 30 May 2026
- Published by Springer Nature in Evolving Systems
- Vol. 17 (3)
Chaotic aquila optimizer enhanced with elite opposite-based learning and variable search strategies
- 30 May 2026
- Published by Springer Nature in Evolving Systems
- Vol. 17 (3)
Chaotic aquila optimizer enhanced with elite opposite-based learning and variable search strategies
- 1 December 2025
- Published by Konya Muhendislik Bilimleri Dergisi in Konya Journal of Engineering Sciences
- Vol. 13 (4) , 1137-1157
THE STUDY OF THE HYPERPARAMETER OPTIMIZATION ANALYSIS IN THE CONVOLUTIONAL NEURAL NETWORKS MODEL WITH STARFISH OPTIMIZATION ALGORITHM
In this study, the hyperparameters of Convolutional Neural Networks (CNNs) have been optimized with the newly proposed Starfish Optimization Algorithm (SFOA) in recent years. CNN has complex hyperparameters due to its structure. In the literature, the values of hyperparameters are mostly tried to be determined with combinatorial methods. The success of metaheuristic algorithms in optimizing the variables of different problems has inspired this study. Thus, four different numbers of channel values (8, 16, 32, and 64), five different kernel size values (1×1, 3×3, 5×5, 7×7, and 9×9), four different batch size values (32, 64, 128, and 256), twenty different values randomly generated between 0 and 0.05 for the learning rate, three different optimizer types (sgdm, adam, and rmsprop), and four different epoch values (5, 10, 15, and 20), which are the most critical hyperparameters in CNN, have been determined. A 6-dimensional solution space was determined with SFOA, and these hyperparameter values were placed in discretely defined dimensions. SFOA tried to determine the most appropriate hyperparameter values for the CNN model in each iteration. In this study, two different image datasets (MNIST and Kuzushiji-MNIST) were selected for CNN classification. Due to the hyperparameter optimization carried out with the SFOA algorithm, an accuracy of 99.52% for the MNIST dataset and 97.91% for the Kuzushiji-MNIST dataset was achieved. Comparisons with existing literature demonstrate that the proposed model showcases successful and competitive performance. Finally, the proposed CNN models are evaluated on a different image dataset, EMNIST (Extended MNIST). EMNIST is a more comprehensive version of MNIST developed for classifying handwritten letters and numbers. The accuracy results on the EMNIST dataset were 88.65% (the proposed CNN model with similar hyperparameter settings as MNIST) and 88.73% (the proposed CNN model with similar hyperparameter settings as Kuzushiji-MNIST), respectively. Additionally, hyperparameters for the EMNIST dataset were determined using SFOA, achieving an accuracy of 88.71%. Analyzing the hyperparameters of three different CNN models, it was observed that similar optimizer types, epoch numbers, kernel sizes, and channel numbers were preferred. This demonstrates that SFOA can produce reliable and effective settings across different datasets.
- 1 December 2025
- Published by Konya Muhendislik Bilimleri Dergisi in Konya Journal of Engineering Sciences
- Vol. 13 (4) , 1137-1157
THE STUDY OF THE HYPERPARAMETER OPTIMIZATION ANALYSIS IN THE CONVOLUTIONAL NEURAL NETWORKS MODEL WITH STARFISH OPTIMIZATION ALGORITHM
In this study, the hyperparameters of Convolutional Neural Networks (CNNs) have been optimized with the newly proposed Starfish Optimization Algorithm (SFOA) in recent years. CNN has complex hyperparameters due to its structure. In the literature, the values of hyperparameters are mostly tried to be determined with combinatorial methods. The success of metaheuristic algorithms in optimizing the variables of different problems has inspired this study. Thus, four different numbers of channel values (8, 16, 32, and 64), five different kernel size values (1×1, 3×3, 5×5, 7×7, and 9×9), four different batch size values (32, 64, 128, and 256), twenty different values randomly generated between 0 and 0.05 for the learning rate, three different optimizer types (sgdm, adam, and rmsprop), and four different epoch values (5, 10, 15, and 20), which are the most critical hyperparameters in CNN, have been determined. A 6-dimensional solution space was determined with SFOA, and these hyperparameter values were placed in discretely defined dimensions. SFOA tried to determine the most appropriate hyperparameter values for the CNN model in each iteration. In this study, two different image datasets (MNIST and Kuzushiji-MNIST) were selected for CNN classification. Due to the hyperparameter optimization carried out with the SFOA algorithm, an accuracy of 99.52% for the MNIST dataset and 97.91% for the Kuzushiji-MNIST dataset was achieved. Comparisons with existing literature demonstrate that the proposed model showcases successful and competitive performance. Finally, the proposed CNN models are evaluated on a different image dataset, EMNIST (Extended MNIST). EMNIST is a more comprehensive version of MNIST developed for classifying handwritten letters and numbers. The accuracy results on the EMNIST dataset were 88.65% (the proposed CNN model with similar hyperparameter settings as MNIST) and 88.73% (the proposed CNN model with similar hyperparameter settings as Kuzushiji-MNIST), respectively. Additionally, hyperparameters for the EMNIST dataset were determined using SFOA, achieving an accuracy of 88.71%. Analyzing the hyperparameters of three different CNN models, it was observed that similar optimizer types, epoch numbers, kernel sizes, and channel numbers were preferred. This demonstrates that SFOA can produce reliable and effective settings across different datasets.
- Open Access
- 31 August 2025
- Published by Plusbase Akademi Organizasyon ve Danismanlik Ltd.Sti. in Intelligent Methods in Engineering Sciences
- Vol. 4 (2) , 29-37
Binary Aquila Optimizer with Taper-Shaped Transfer Function: An Application for Merkle-Hellman Knapsack Cryptosystems
Metaheuristic algorithms are powerful methods used to solve large and complex optimization problems. Thanks to their flexibility, they provide effective results in various fields and also have an important place in security applications such as the Merkle-Hellman Knapsack Crypto System. Aquila Optimizer is an optimization algorithm inspired by the hunting behavior of aquilas. It provides fast and effective solutions to complex problems. In this study, Aquila Optimizer is discretized using taper-shaped transfer functions. Taper-shaped transfer functions help the algorithm obtain more precise and effective results by increasing its performance. Four BinAO versions obtained in this way were tested on the Merkle-Hellman Knapsack Cryptosystem. In the tests conducted for the cryptanalysis of "CAT" and "MACRO" messages, the version achieved more successful results. Additionally, tests conducted with the algorithms in the literature clearly showed that the proposed algorithm is successful and effective.
- 31 August 2025
- Published by Plusbase Akademi Organizasyon ve Danismanlik Ltd.Sti. in Intelligent Methods in Engineering Sciences
- Vol. 4 (2) , 29-37
Open AccessBinary Aquila Optimizer with Taper-Shaped Transfer Function: An Application for Merkle-Hellman Knapsack Cryptosystems
Metaheuristic algorithms are powerful methods used to solve large and complex optimization problems. Thanks to their flexibility, they provide effective results in various fields and also have an important place in security applications such as the Merkle-Hellman Knapsack Crypto System. Aquila Optimizer is an optimization algorithm inspired by the hunting behavior of aquilas. It provides fast and effective solutions to complex problems. In this study, Aquila Optimizer is discretized using taper-shaped transfer functions. Taper-shaped transfer functions help the algorithm obtain more precise and effective results by increasing its performance. Four BinAO versions obtained in this way were tested on the Merkle-Hellman Knapsack Cryptosystem. In the tests conducted for the cryptanalysis of "CAT" and "MACRO" messages, the version achieved more successful results. Additionally, tests conducted with the algorithms in the literature clearly showed that the proposed algorithm is successful and effective.
- Open Access
- 19 August 2025
- Published by Elsevier in Alexandria Engineering Journal
- Vol. 129, 1061-1108
A study on gender detection using multiple classifiers on voice data
- 19 August 2025
- Published by Elsevier in Alexandria Engineering Journal
- Vol. 129, 1061-1108
Open AccessA study on gender detection using multiple classifiers on voice data
- Open Access
- 24 April 2025
- Published by Eskisehir Osmangazi Universitesi Muhendislik ve Mimarlik Fakultesi Dergisi in Eskişehir Osmangazi Üniversitesi Mühendislik ve Mimarlık Fakültesi Dergisi
- Vol. 33 (1) , 1695-1711
BINARY HONEY BADGER ALGORITHM ENHANCED WITH TIME-VARYING SIGMOID TRANSFER FUNCTION AND CROSSOVER STRATEGY
Modeling the foraging behavior of honey badgers, the Honey Badger Algorithm (HBA) is a recently proposed metaheuristic algorithm. In this study, a binary version of this algorithm that was proposed for solving continuous optimization problems was developed. The S-shaped transfer function and crossover strategy were used to transform the continuous algorithm into a binary algorithm. Eight S-shaped transfer functions with constant and time-varying features were used, and the most successful function was determined. Additionally, the effect of time-varying transfer functions was examined. Three strategies, single-point, two-point, and uniform, were applied as crossover strategies, and the uniform strategy, which was more successful than others, was integrated into the algorithm. The binary algorithm (BinHBA) developed in this way was tested on a total of twenty-seven knapsack problems, fifteen small-scale and twelve large-scale. Statistical tests were employed to analyze the results and compare them with methods found in the existing literature. The results showed that the proposed BinHBA for binary optimization problems is effective and preferable.
- 24 April 2025
- Published by Eskisehir Osmangazi Universitesi Muhendislik ve Mimarlik Fakultesi Dergisi in Eskişehir Osmangazi Üniversitesi Mühendislik ve Mimarlık Fakültesi Dergisi
- Vol. 33 (1) , 1695-1711
Open AccessBINARY HONEY BADGER ALGORITHM ENHANCED WITH TIME-VARYING SIGMOID TRANSFER FUNCTION AND CROSSOVER STRATEGY
Modeling the foraging behavior of honey badgers, the Honey Badger Algorithm (HBA) is a recently proposed metaheuristic algorithm. In this study, a binary version of this algorithm that was proposed for solving continuous optimization problems was developed. The S-shaped transfer function and crossover strategy were used to transform the continuous algorithm into a binary algorithm. Eight S-shaped transfer functions with constant and time-varying features were used, and the most successful function was determined. Additionally, the effect of time-varying transfer functions was examined. Three strategies, single-point, two-point, and uniform, were applied as crossover strategies, and the uniform strategy, which was more successful than others, was integrated into the algorithm. The binary algorithm (BinHBA) developed in this way was tested on a total of twenty-seven knapsack problems, fifteen small-scale and twelve large-scale. Statistical tests were employed to analyze the results and compare them with methods found in the existing literature. The results showed that the proposed BinHBA for binary optimization problems is effective and preferable.
- Open Access
- 28 March 2024
- Published by Anadolu Universitesi Bilim ve Teknoloji Dergisi-A: Uygulamali Bilimler ve Muhendislik in Eskişehir Technical University Journal of Science and Technology A - Applied Sciences and Engineering
- Vol. 25 (1) , 78-98
PERFORMANCE EVALUATIONS OF THE MANTA RAY FORAGING OPTIMIZATION ALGORITHM IN REAL-WORLD CONSTRAINED OPTIMIZATION PROBLEMS
Metaheuristic algorithms are often preferred for solving constrained engineering design optimization problems. The most important reason for choosing these algorithms is that they guarantee a satisfactory response within a reasonable time. The swarm intelligence-based manta ray foraging optimization algorithm (MRFO) is a metaheuristic algorithm proposed to solve engineering applications. In this study, the performance of MRFO is evaluated on 19 mechanical engineering optimization problems in the CEC2020 real-world constrained optimization problem suite. In order to increase the MRFO performance, three modifications are made to the algorithm; in this way, the enhanced manta ray foraging optimization (EMRFO) algorithm is proposed. The effects of the modifications made are analyzed and interpreted separately. Its performance has been compared with the algorithms in the literature, and it has been shown that EMRFO is a successful and preferable algorithm for this problem suite.
- 28 March 2024
- Published by Anadolu Universitesi Bilim ve Teknoloji Dergisi-A: Uygulamali Bilimler ve Muhendislik in Eskişehir Technical University Journal of Science and Technology A - Applied Sciences and Engineering
- Vol. 25 (1) , 78-98
Open AccessPERFORMANCE EVALUATIONS OF THE MANTA RAY FORAGING OPTIMIZATION ALGORITHM IN REAL-WORLD CONSTRAINED OPTIMIZATION PROBLEMS
Metaheuristic algorithms are often preferred for solving constrained engineering design optimization problems. The most important reason for choosing these algorithms is that they guarantee a satisfactory response within a reasonable time. The swarm intelligence-based manta ray foraging optimization algorithm (MRFO) is a metaheuristic algorithm proposed to solve engineering applications. In this study, the performance of MRFO is evaluated on 19 mechanical engineering optimization problems in the CEC2020 real-world constrained optimization problem suite. In order to increase the MRFO performance, three modifications are made to the algorithm; in this way, the enhanced manta ray foraging optimization (EMRFO) algorithm is proposed. The effects of the modifications made are analyzed and interpreted separately. Its performance has been compared with the algorithms in the literature, and it has been shown that EMRFO is a successful and preferable algorithm for this problem suite.
- 10 December 2023
- Published by Springer Nature in Neural Computing and Applications
- Vol. 36 (8) , 4151-4177
A new binary arithmetic optimization algorithm for uncapacitated facility location problem
- 10 December 2023
- Published by Springer Nature in Neural Computing and Applications
- Vol. 36 (8) , 4151-4177
A new binary arithmetic optimization algorithm for uncapacitated facility location problem
- 24 November 2023
- Published by Springer Nature in Neural Computing and Applications
- Vol. 36 (6) , 2797-2834
A new binary coati optimization algorithm for binary optimization problems
- 24 November 2023
- Published by Springer Nature in Neural Computing and Applications
- Vol. 36 (6) , 2797-2834
A new binary coati optimization algorithm for binary optimization problems
- 28 June 2023
- Published by Springer Nature in Neural Processing Letters
- Vol. 55 (8) , 10131-10199
Enhanced Coati Optimization Algorithm for Big Data Optimization Problem
- 28 June 2023
- Published by Springer Nature in Neural Processing Letters
- Vol. 55 (8) , 10131-10199
Enhanced Coati Optimization Algorithm for Big Data Optimization Problem
- Open Access
- 28 April 2023
- Published by Konya Muhendislik Bilimleri Dergisi in Konya Journal of Engineering Sciences
- Vol. 11 (3) , 610-634
A DISCRETE PARTICLE SWARM ALGORITHM WITH SYMMETRY METHODS FOR DISCRETE OPTIMIZATION PROBLEMS
Particle Swarm Optimization (PSO) is a commonly used optimization to solve many problems. The PSO, which is developed for continuous optimization, is updated to solve discrete problems and Discrete PSO (DPSO) is obtained in this study. With DPSO, the Traveling Salesman Problem (TSP), which is well-known in the literature as a discrete problem, is solved. In order to improve the results, the swap method, the shift method, and the symmetry method are added to DPSO. The symmetry method is a new and successful method. The variations of the DPSO occurred according to the selected method type (DPSO1 (swap method), DPSO2 (shift method), DPSO3 (swap and shift methods), DPSO4 (symmetry method), DPSO5 (swap, shift, and symmetry methods), DPSO6 (swap, shift, symmetry, and 2-opt methods)). The effect of each method on the performance of the DPSO has been studied in detail. To demonstrate the success of the variations of the DPSO, the results are additionally compared with many well-known and new discrete algorithms in the literature. The results showed that the performance of DPSO has improved with the symmetry method and it has achieved better results than the discrete heuristic algorithms recently proposed in the literature.
- 28 April 2023
- Published by Konya Muhendislik Bilimleri Dergisi in Konya Journal of Engineering Sciences
- Vol. 11 (3) , 610-634
Open AccessA DISCRETE PARTICLE SWARM ALGORITHM WITH SYMMETRY METHODS FOR DISCRETE OPTIMIZATION PROBLEMS
Particle Swarm Optimization (PSO) is a commonly used optimization to solve many problems. The PSO, which is developed for continuous optimization, is updated to solve discrete problems and Discrete PSO (DPSO) is obtained in this study. With DPSO, the Traveling Salesman Problem (TSP), which is well-known in the literature as a discrete problem, is solved. In order to improve the results, the swap method, the shift method, and the symmetry method are added to DPSO. The symmetry method is a new and successful method. The variations of the DPSO occurred according to the selected method type (DPSO1 (swap method), DPSO2 (shift method), DPSO3 (swap and shift methods), DPSO4 (symmetry method), DPSO5 (swap, shift, and symmetry methods), DPSO6 (swap, shift, symmetry, and 2-opt methods)). The effect of each method on the performance of the DPSO has been studied in detail. To demonstrate the success of the variations of the DPSO, the results are additionally compared with many well-known and new discrete algorithms in the literature. The results showed that the performance of DPSO has improved with the symmetry method and it has achieved better results than the discrete heuristic algorithms recently proposed in the literature.