Sürdürülebilir Kalkınma Amaçları
Yazarlar
- Tapkın, Serkan(2)
- Abdioglu, Murat(1)
- Aida-zade, K.(1)
- Akpolat, Hacer(1)
- Akpolat, Mehmet Zahid(1)
- Aksungur, Nurhak(1)
- Aliev, T.(1)
- Altundaş, Necip(1)
- Aras, Selim(1)
- Başar, Mehmet Sinan(1)
- Bilen, Zeynep(1)
- Bozkurt, Ferhat(1)
- Dayi, Faruk(1)
- Demirtaş, Ali(1)
- Engin, Mustafa Alptekin(1)
- Erdogdu, Aylin(1)
- Esmer, Yusuf(1)
- Eygu, Hakan(1)
- Gangal, Ali(1)
- Ganji, Farshad(1)
- Kara, Abdulkadir(1)
- Kara, Salih(1)
- Kara, Zeynep Avinc(1)
- Korkut, Ercan(1)
- Kul, Sinan(1)
- Küçük, Furkan(1)
- Li, Yongsheng(1)
- Mammadova, G.(1)
- Nabiyev, Vasif(1)
- Ozkat, Erkan Caner(1)
- Ozturk, U. Kemal(1)
- Pan, Yue(1)
- Senyer Yapici, İrem(1)
- Song, Xieqing(1)
- Teke, Çağatay(1)
- Tercan, Emre(1)
- Turhal, Uğur(1)
- Türker, Abdussamet(1)
- Uzun Arslan, Rukiye(1)
- Yağanoğlu, Mete(1)
Yayınlar
- MakaleAçık Erişim2026Scopus SJRQ1A Hybrid Deep Learning–Fuzzy–Genetic Framework for Climate-Resilient Agricultural Investment and Resource Allocation Under Carbon Market Uncertainty
Erdogdu, Aylin; Dayi, Faruk; Yildiz, Ferah; Esmer, Yusuf; Ganji, Farshad
Climate variability, environmental uncertainty, and carbon-market dynamics increasingly challenge agricultural investment and resource allocation decisions worldwide. This study proposes an integrated hybrid decision-support framework combining Long Short-Term Memory (LSTM) deep learning, Interval Type-2 Fuzzy Logic Systems, and Genetic Algorithms to support climate-resilient agricultural investment analysis under uncertainty. The framework integrates predictive modeling, uncertainty representation, and multi-objective optimization within a unified computational architecture. The empirical analysis was conducted using agricultural, climate, and carbon-market datasets covering Europe, Asia, and Africa over the 2010–2025 period. Agricultural productivity indicators, commodity price variables, climate-risk parameters, and carbon-market data were integrated into the modeling process. LSTM models were employed to analyze temporal agricultural and climate-related dynamics, while Interval Type-2 fuzzy systems were used to represent ambiguity associated with environmental and market uncertainty. Genetic Algorithms were subsequently applied to optimize investment allocation under conflicting objectives related to profitability, sustainability, and risk. The findings suggest that the proposed hybrid framework may improve adaptive investment evaluation and optimization performance under uncertain climate conditions relative to standalone computational approaches within the scope of the analyzed datasets. The results further highlight the importance of integrating predictive analytics, uncertainty modeling, and sustainability-oriented optimization within agricultural decision-support systems. However, the framework should be interpreted as a climate-resilient decision-support architecture rather than a universally deterministic forecasting mechanism. Overall, the study contributes to the emerging literature on agricultural sustainability and climate-resilient investment by presenting a transparent and uncertainty-aware computational framework under evolving environmental and carbon-market conditions.
AtıfScopus1WoS1OpenAlex2
İncele - MakaleYalnız künye2026Scopus SJRQ2Leakage-safe benchmarking of conventional chemometric and deep learning models for food NIR spectroscopy regression
Akpolat, Hacer; Akpolat, Mehmet Zahid
Near-infrared (NIR) spectroscopy is widely used in food systems for rapid, non-destructive estimation of compositional and quality attributes. Although machine-learning and deep-learning models are increasingly reported for spectral regression, practical model selection is often hindered by heterogeneous validation setups and the risk of information leakage during preprocessing and tuning. In this study, we provide a leakage-safe benchmark across nine food-related regression tasks spanning different sample sizes and spectral dimensionalities. Conventional chemometric and regression baselines (partial least squares regression, ridge regression, and support vector regression) are compared with lightweight deep-learning baselines (a shallow artificial neural network and a one-dimensional convolutional neural network) under a unified leakage-safe evaluation framework combining nested cross-validation with acomplementary fixed held-out test-set assessment, fold-wise preprocessing, and repeated training for stochastic models. Performance is summarized primarily using prediction error (RMSE in nested cross-validation and RMSEP on the held-out test set), while the coefficient of determination is retained as a secondary descriptive measure, complemented by cross-validation and fixed held-out-test learning-curve analyses together with model-complexity analysis. Across both validation analyses, ridge regression achieved the largest number of numerical task-level wins, while partial least squares regression and support vector regression remained competitive across multiple tasks. Paired randomization tests following van der Voet further showed that some small numerical differences between models were not statistically distinguishable. Artificial neural network and one-dimensional convolutional neural network baselines did not provide a reliable generalization advantage and showed greater variability in some data-limited settings. From a food-measurement perspective, these findings indicate that well-validated conventional models remain a strong default for routine NIR calibration workflows and non-destructive quality assessment. Deep learning appears most justified when calibration settings provide larger datasets, stronger experimental control, or spectroscopy-specific inductive bias that can translate added model complexity into practical measurement gains. Overall, this study serves as a controlled benchmark and reference for model selection under realistic food-spectroscopy data regimes.
AtıfOpenAlex1İncele - MakaleAçık Erişim2026Scopus SJRQ2Deep Learning-Based Classification of Common Lung Sounds via Auto-Detected Respiratory Cycles
Engin, Mustafa Alptekin; Uzun Arslan, Rukiye; Senyer Yapici, İrem; Aras, Selim; Gangal, Ali
Chronic respiratory diseases, the third leading cause of mortality on a global scale, can be diagnosed at an early stage through non-invasive auscultation. However, effective manual differentiation of lung sounds (LSs) requires not only sharp auditory skills but also significant clinical experience. With technological advancements, artificial intelligence (AI) has demonstrated the capability to distinguish LSs with accuracy comparable to or surpassing that of human experts. This study broadly compares the methods used in AI-based LSs classification. Firstly, respiratory cycles—consisting of inhalation and exhalation parts in LSs of different lengths depending on individual variability, obtained and labelled under expert guidance—were automatically detected using a series of signal processing procedures and a database was obtained in this way. This database of common LSs was then classified using various time-frequency representations such as spectrograms, scalograms, Mel-spectrograms and gammatonegrams for comparison. The utilisation of proven, convolutional neural network (CNN)-based pre-trained models through the application of transfer learning facilitated the comparison, thereby enabling the acquisition of the features to be employed in the classification process. The performances of CNN, CNN and Long Short-Term Memory (LSTM) hybrid architecture and support vector machine methods were compared in the classification process. When the spectral structure of gammatonegrams, which capture the spectral structure of signals in the low-frequency range with high fidelity and their noise-resistant structures, is combined with a CNN architecture, the best classification accuracy of 97.3% ± 1.9 is obtained.
AtıfScopus2WoS2OpenAlex2
İncele - MakaleYayıncıda açık erişim2025Answer-based and reference-based BERT models for automatic scoring of Turkish short answers: The decisive role of task complexity
Kara, Abdulkadir; Kara, Zeynep Avinc; Yildirim, Serkan
In measurement and evaluation processes, natural language responses are often avoided due to time, workload, and reliability concerns. However, the increasing popularity of automatic short-answer grading studies for natural language responses means such answers can now be measured more quickly and reliably. This study aims to build models for predicting automatic short answer scores using the pre-trained BERT deep learning language model and to reveal their effectiveness. For this purpose, two different score prediction models were created using an answerbased approach that aligns student answers with expert judgements and a referencebased approach that matches student answers with reference answers. The dataset includes answers from 246 Physics department students responding to 4 physicsrelated questions. The performance of these models was evaluated on four physics questions representing varying levels of cognitive complexity, using Cohen's Kappa for statistical comparison of agreement with expert scores. Our findings reveal a clear interaction between model architecture and task complexity. The answer-based model was unequivocally superior for the most complex, multi-class task, effectively capturing diverse, nuanced responses. Conversely, the reference-based model demonstrated a statistically significant advantage for a well-defined, mediumcomplexity binary task. This study concludes that the optimal model for ASAG in Turkish is contingent on the cognitive demands of the assessment task, suggesting that a onesize-fits-all solution may not be the most effective approach. This provides a critical framework for practitioners, demonstrating not only that effective models are feasible for complex languages, but that their selection must be guided by task complexity.
AtıfWoS1OpenAlex2
İncele - MakaleAçık Erişim2025Scopus SJRQ2Development of a Clinical Decision Support System Using Artificial Intelligence Methods for Liver Transplant Centers
Yağanoğlu, Mete; Öztürk, Gürkan; Bozkurt, Ferhat; Bilen, Zeynep; Yetiş Demir, Zühal; Kul, Sinan; Şimşek, Emrah; Kara, Salih; Eygu, Hakan; Altundaş, Necip
The objective of this study is to utilize artificial intelligence techniques for the diagnosis of complications and diseases that may arise after liver transplantation, as well as for the identification of patients in need of transplantation. To achieve this, an interface was developed to collect patient information from Atatürk University Research Hospital, specifically focusing on individuals who have undergone liver transplantation. The collected data were subsequently entered into a comprehensive database. Additionally, relevant patient information was obtained through the hospital’s information processing system, which was used to create a data pool. The classification of data was based on four dependent variables, namely, the presence or absence of death (“exitus”), recurrence location, tumor recurrence, and cause of death. Techniques such as Principal Component Analysis and Linear Discriminant Analysis (LDA) were employed to enhance the performance of the models. Among the various methods employed, the LDA method consistently yielded superior results in terms of accuracy during k-fold cross-validation. Following k-fold cross-validation, the model achieved the highest accuracy of 98% for the dependent variable “exitus”. For the dependent variable “recurrence location”, the highest accuracy obtained after k-fold cross-validation was 91%. Furthermore, the highest accuracy of 99% was achieved for both the dependent variables “tumor recurrence” and “cause of death” after k-fold cross-validation.
AtıfScopus2WoS2OpenAlex2İncele - BildiriAçık Erişim2025Scopus SJRQ4Lightweight Hybrid CNN Model for Face Presentation Attack Detection
Turhal, Uğur; Yilmaz, Asuman Günay; Nabiyev, Vasif; Mammadova, G.; Aliev, T.; Aida-zade, K.
Today, face recognition systems are widely used in many areas that require biometric-based verification, especially because they are contactless and require low user cooperation. Despite their ease of implementation, these systems are vulnerable to attacks. Especially the increasing use of social media, makes it easier to spoof face recognition systems. Therefore, it is very important to develop new methods for Face Presentation Attack (FPA) detection. In this study, FPA detection performances of the lightweight Convolutional Neural Network (CNN) models -MobileNetV3-Small, ShuffleNet, SqueezeNet- and lightweight hybrid CNN models designed with combinations of these networks were investigated. In the experiments, the change in FPA detection performance was examined by giving the cropped face regions and the whole images as input to the networks. Using the whole image instead of the cropped face region significantly improved the FPA detection performance of the CNN models. These results showed that non-facial areas carry important information in FPA detection. Also, the FPA detection performance of the proposed lightweight hybrid CNN models were better than the single use of the networks for both inputs. 0.17% HTER performance was achieved with the hybrid CNN model built from the combination of ShuffleNet + SqueezeNet on the Replay-Attack dataset.
- MakaleAçık Erişim2024Scopus SJRQ4PREDICTION OF TURKEY'S COTTON SOCK EXPORTS TO GERMANY USING DEEP LEARNING APPROACH; [TÜRKİYE'NİN ALMANYA'YA PAMUKLU ÇORAP İHRACATININ DERİN ÖĞRENME YAKLAŞIMI İLE TAHMİNİ]
Özbek, Ahmet; Teke, Çağatay
Cotton socks are a strategic export product for Turkey. Therefore, the aim of this study is to forecast Turkey's exports to Germany, the world's largest cotton socks market. In order to achieve this objective, the determinants of exports were identified by analysing the literature. Then, expert opinion was sought to determine the importance of these factors for Turkey's cotton socks exports to Germany. Using the deep learning model created from the factors determined as a result of the expert opinion, the prediction of the export of Turkish socks to Germany was realised. A success rate of 96% was achieved with the prediction.
AtıfScopus1OpenAlex1TR Dizin1İncele - MakaleAçık Erişim2024Scopus SJRQ3Machine learning driven optimization and parameter selection of multi-surface HTS Maglev
Ozkat, Erkan Caner; Abdioglu, Murat; Ozturk, U. Kemal
This research aims to tackle the challenges posed by precise force measurement for high temperature superconducting (HTS) Maglev systems, including mechanical constraints, step motor limitations, and sensor resolutions. For this aim, six machine learning (ML) models namely Support Vector Machine (SVM), Gaussian Process Regression (GPR), Extreme Gradient Boosting (XGB), Long Short-Term Memory (LSTM), Extreme Machine Learning (EML), and Convolutional Neural Network (CNN) were developed to predict levitation force (Fz) and lateral force (Fx) based on process parameters including permanent magnet width (PMW), field cooling height (FCH), the movement in the z-axis (vertical distance), and the movement in the x-axis (lateral distance). Among six ML models, CNN emerged as the most accurate model, demonstrating smaller root mean square deviation (RMSD) without compromising correlation coefficients. Furthermore, an innovative process window approach was introduced to select process parameters that simultaneously meet the minimum value of Fz and maximum value of Fx, named β1 and β2, set at 90 N and 0 N, respectively. Within this window, PMW of 30 mm and z values less than 10 mm were found to be consistent for all FCH and x values. The novelty of this study is to formulate the optimisation problem in HTS Maglev using the developed ML model by addressing two specific objectives one of which focuses on maximizing Fz while ensuring Fx remains within a defined tolerance (β3), representing the minimum allowable ratio of the levitation force to the total force, and the second problem aims to maximize Fz while obtaining zero Fx. The optimum PMW, FCH, x, and z values were obtained at 30 mm, 30 mm, 4 mm and 5 mm, corresponding to Fz and Fx values of 224.2 N and -53.8 N for option 1. As for option 2, the process parameters were obtained as 28.6 mm, 25.9 mm, 0 mm, and 5 mm, corresponding to Fz and Fx values of 194.2 N and 0 N. It was obtained both experimentally and by the optimization that Fz reaches close its maximum as the Fx gains attractive character. Hence, it is expected that the outcomes of this study will significantly benefit the design of HTS Maglev systems and find valuable applications across various transportation engineering projects.
AtıfScopus21WoS18İncele - MakaleAçık Erişim2024Scopus SJRQ1Energy-driven TBM health status estimation with a hybrid deep learning approach
Li, Yongsheng; Zhang, Limao; Pan, Yue; Tapkın, Serkan; Song, Xieqing
This study provides a hybrid deep learning approach that enables the researchers to estimate the Tunnel Boring Machine (TBM) health status during excavation process using energy consumption data. By analyzing the energy forms and paths in the tunneling process, the key energy factors' consumption performance behavior affecting TBM safety is identified. The Ensemble Empirical Mode Decomposition (EEMD) method and a Convolution neural network-Long short term memory-Depth Neural Network (CNN-LSTM-DNN, CLDNN) model are integrated for Health Performance Parameter (HPP) prediction. A classification method is constructed by the Generative Adversarial Imputation Networks (GAIN) model for health status estimation. The SHapley Additive exPlanations (SHAP) method is performed to quantify the relationships between energy consumption and system health status. A tunneling case in Singapore is used to test the effectiveness of the developed methodology. It is found that (1) the proposed method provides TBM cutter wear and cooling oil prediction with a R-square (R2) of 0.8591 and 0.8727, respectively, and (2) classification results indicate that the estimation F1-score of 0.9863 can be achieved based on health status evaluation indicator (HSEI). The novelty of the proposed approach lies in its capabilities of (1) fully extracting and utilizing the energy consumption information in every regard to predict TBM health performance; (2) providing a good one-step-ahead health status estimation for the TBM excavation process. Therefore, this study proposes a feasible direction for the application of energy consumption data in TBM health status estimation.
AtıfScopus17WoS15OpenAlex17İncele - MakaleAçık Erişim2023Scopus SJRQ2Computational Intelligence Based Point of Interest Detection by Video Surveillance Implementations
Tercan, Emre; Tapkın, Serkan; Küçük, Furkan; Demirtaş, Ali; Özbayoğlu, Ahmet; Türker, Abdussamet
Latest advancement of the computer vision literature and Convolutional Neural Networks (CNN) reveal many opportunities that are being actively used in various research areas. One of the most important examples for these areas is autonomous vehicles and mapping systems. Point of interest detection is a rising field within autonomous video tracking and autonomous mapping systems. Within the last few years, the number of implementations and research papers started rising due to the advancements in the new deep learning systems. In this paper, our aim is to survey the existing studies implemented on point of interest detection systems that focus on objects on the road (like lanes, road marks), or objects on the roadside (like road signs, restaurants or temporary establishments) so that they can be used for autonomous vehicles and automatic mapping systems. Meanwhile, the roadside point of interest detection problem has been addressed from a transportation industry perspective. At the same time, a deep learning based point of interest detection model based on roadside gas station identification will be introduced as proof of the anticipated concept. Instead of using an internet connection for point of interest retrieval, the proposed model has the capability to work offline for more robustness. A variety of models have been analysed and their detection speed and accuracy performances are compared. Our preliminary results show that it is possible to develop a model achieving a satisfactory real-time performance that can be embedded into autonomous cars such that streaming video analysis and point of interest detection might be achievable in actual utilisation for future implementations.
AtıfScopus1OpenAlex1İncele