SŞSevil Şentürk

Sevil Şentürk

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Sevil Şentürk has a total of 18 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • A new method for modelling the stochastic differential equations

    by Nihal İnCe, Sevil Şentürk

    This study presents a novel approach to estimate the probability density function of solutions to stochastic differential equations using generalized entropy optimization methods. Unlike traditional methods such as the Fokker–Planck–Kolmogorov equation, the proposed generalized entropy optimization methods framework accommodates cases where the distribution of the solution deviates from standard statistical forms. The method integrates the Euler–Maruyama scheme to generate multiple trajectories, producing random variables $\hat{X}(t)$ for each time $t$. The performance of method is evaluated through a comprehensive simulation study, in which it is compared with existing techniques under various parameter settings. Both generalized MaxEnt and MinxEnt distributions are applied, with results indicating that generalized MinxEnt distributions offer superior adaptability and accuracy. Visual and statistical comparisons confirm the theoretical validity and practical efficiency of the method. This framework not only provides a flexible alternative for probability density function estimation in stochastic differential equation modeling but also opens pathways for applications in fuzzy stochastic differential equation systems.

  • Open Access

    Optuna-Optimized Pythagorean Fuzzy Deep Neural Network: A Novel Framework for Uncertainty-Aware Image Classification

    by Asli Kaya Karakutuk, Ozer Ozdemir, Sevil Senturk

    By using Geographic Information Systems, satellite imagery from remote sensing techniques provides quantitative and qualitative data about Earth’s natural and human elements. However, the direct use of raw imagery may prevent the accurate identification of the spectral and temporal characteristics of the target objects. To obtain meaningful results from these data, the object and surface features in the image must be classified correctly. In this context, this study develops a new deep learning approach that includes hyperparameter optimization that considers uncertainty factors when classifying satellite imagery. In the proposed approach, a hybrid architecture called CNN-Pythagorean Fuzzy Deep Neural Network (PFDNN) is developed by combining the ability of convolutional neural networks (CNN) to reveal expressive features with the ability of Pythagorean fuzzy set (PFS) theory to predict uncertainty. In addition, to further improve the model’s success, the hyperparameters are automatically optimized using Optuna. In the experiments conducted on the EuroSAT RGB dataset, the CNN+PFDNN+Optuna model achieved 0.9696 ± 0.0037 accuracy and a macro-AUC value of 0.9983, outperforming other methods such as DNN, FDNN, PFDNN and VGG16+PFDNN. Including the Pythagorean fuzzy layer in the system provided about 13.05% higher accuracy than conventional fuzzy systems. These results show that integrating the Pythagorean fuzzy set approach into deep learning models contributes to more effective management of uncertainties in remote sensing data and that hyperparameter optimization significantly impacts model performance.


  • USING CIRCULAR INTUITIONISTIC FUZZY TOPSIS TO ELIMINATE THE HESITANCY IN THE FIELD OF AGRICULTURAL BIOLOGY

    by Zeynep Gökkuş, Sevil Şentürk, Fırat Alatürk, Baboo Ali

    Circular intuitionistic fuzzy sets (CIFSs) were introduced by Atanassov (2020) as a new extension of intuitionistic fuzzy sets. C-IFSs are represented by a circle of each element that is characterized by degrees of membership and non-membership. In a decision-making process based on experimental data, although uncertainty is low, hesitation can be high. In such cases, the decision-making process is affected by the decision-makers as well as the criteria. Therefore, there is a need to evaluate the expertise of decision-makers within the decision-making process. In Circular Intuitionistic Fuzzy Sets, where hesitation is represented, an approach is proposed for calculating decision-maker weights with the Technique for Order of Preferences by Similarity to Ideal Solution (TOPSIS), which is a multi-criteria decision-making method to eliminate hesitation. In this study, circular intuitionistic fuzzy sets were implemented into the TOPSIS method. The problem was handled from two different perspectives while creating the decision matrix. Sensitivity analyses were performed for both applications. These sensitivity analyses were carried out to examine the change in the ranking of the alternatives when the optimistic or pessimistic approaches of the decision makers, criterion weights, and the decision maker importance weights changed, respectively. In addition, intuitionistic and Pythagorean fuzzy TOPSIS methods were applied and presented as comparative analyses. According to the results obtained, the proposed approaches were satisfactory in eliminating the hesitation.


  • Open Access

    Sezgisel Bulanık Tek-Yönlü ANOVA Modeli ve Bir Uygulama

    by Zeynep Gökkuş, Sevil Şentürk

    Varyans analizi (ANOVA) modelleri ikiden fazla bağımsız grubun ortalamalarının karşılaştırılmasında istatistiksel çalışmalarda kullanılan önemli yöntemlerdir. Uzun yıllardır hemen her alanda, bilimsel araştırmalarda ANOVA modelleri yaygın olarak kullanılmaktadır. Tek-yönlü ANOVA modelinde tek faktör (bağımsız değişken) ve bu faktörün çeşitli düzeyleri ya da denemeleri söz konusu olmaktadır. Burada amaç denemelerin bağımlı değişken üzerindeki etkilerini incelemektir. Belirsizlik altında elde edilen ve bulanık sayılarla temsil edilen gözlemlerin modellenmesi için de farklı bulanık ANOVA modelleri geliştirilmiştir. Literatürde yapılan çalışmalar istatistiksel olarak incelendiğinde, sezgisel bulanık ANOVA modellerine ilişkin çalışmaların az sayıda olduğu görülmüştür. Bu çalışmada sezgisel bulanık tek-yönlü ANOVA modelinin teorik yapısı anlatılmış ve bir örnek uygulama ile açıklanmıştır. Elde edilen sonuçlar karşılaştırma analizi ile desteklenmiş ve yorumlanmıştır.

  • Open Access

    Comparison of Restricted and Unrestricted estimators with a Monte-Carlo Simulation Study

    by Israa Shaltoot, Nihal İnCe, Sevil Şentürk

    This study deals with the problem of multicollinearity in the linear regression model. Restricted and unrestricted parameter estimates are chosen among biased estimators to be studied and compared as two corresponding groups, with the aim of identifying which group gives better parameter estimates in the case of multicollinearity. Estimators' performance is compared according to matrix mean square error and scalar mean square error. Proceeding from this, it has been shown that, in the sense of Scalar Mean Square Error (SMSE), the Restricted Ridge regression (RRR) estimator outperforms all constrained and unconstrained estimators, while the Ridge regression is superior to the unconstrained set of estimators. A real-life application and Monte-Carlo simulation study are conducted to compare the performance of restricted and unrestricted estimators. As a result, it was decided that the most effective estimators are the restricted biased estimators when it comes to the state of multicollinearity.

  • A Multivariate Intuitionistic Fuzzy Grey Model for Forecasting Electricity Consumption

    by Bahadır Yörür, Nihal Erginel, Sevil Şentürk

    Background: Whether in the short, medium or long term, forecasting electricity consumption has always been an essential study area. In the literature, many methods are used for future forecasting and are being improved daily to achieve better results. Objective: The main objective of this study is to make the most accurate long-term electricity consumption forecast, which is the basis for optimal future planning in the energy sector. Electric consumption forecasting is performed regionally since planning at the regional level is essential for more precise planning. Methods: There may be different variables that affect electricity consumption. This study ex-tends the multivariate grey model for electricity consumption prediction to intuitionistic trian-gular fuzzy numbers for nine regions. In the grey model, population, export, and gross domes-tic product variables were used as independent variables, and future predictions for these vari-ables were obtained through the univariate intuitionistic triangular fuzzy grey model. Results: The results of the proposed method are compared with those of the classical univari-ate grey model, univariate intuitionistic triangular fuzzy grey model, and classical multivariate grey model. The results show that the error values of the proposed method are lower. Conclusion: The study contributes to the development of the grey model. More accurate pre-diction results are obtained with the proposed method compared to similar methods


  • Theoretical Structure of Circular Intuitionistic Fuzzy C Control Charts

    by Sevil Şentürk, Zeynep Gökkuş, Gültekin Atalik, Nihal Ince, Asiye Özbek Yazicioğlu

  • Open Access

    THE USE OF COX REGRESSION MODEL IN THE SURVIVAL ANALYSIS FOR LEUKEMIA PATIENTS IN THE REPUBLIC OF YEMEN

    by Elıas Abdullah AL-SAMAI, Sevil Şentürk

    This study aims at analyzing and studying the theoretical and practical importance of the (Cox) regression model in the analysis of survival as well as measuring the most important factors affecting the survival time for patients with leukemia. Moreover, it aims at reaching the expected survival time for patients and creating a life table for patients by using (Cox) regression model. To achieve these goals, real data were taken for (1168) patients with leukemia in the Republic of Yemen in the period from January 2017 to February 2022. The dependent variable, which is the patient’s condition at the end of the period, was determined in addition to the patient’s survival time and eight independent variables were identified. The effect of these variables on the survival time of patients with leukemia was investigated using the SPSS program. The study concluded several results, the most prominent of them are the following: There are no differences in the incidence rate between males and females and the most age group affected by this disease is (40 years and over). Furthermore, it was found that acute lymphoblastic leukemia (ALL) is the most prevalent type among the other types and there is a difference in the risk of death among those who take intravenous chemotherapy and those patients who take oral chemotherapy. Other significant result was found that there is a higher risk of death for non-regular patients in receiving treatment compared to regular patients in receiving treatment. It was found that the most influencing variables on the survival time of patients are (age, marital status, type of disease, the governorate in which the patient lives, regularity in receiving treatment and type of chemotherapy). Through the life table, it is noticed that the greatest risk in the survival time is in the thirty-sixth month, which is the largest among all other periods, as it reached (0.10). Additionally, the median survival time was reached (35.06) months. Finally, the study found that there are differences in the incidence according to the type of disease in terms of the risk of death as acute lymphoblastic leukemia (ALL) is the most prevalent disease among all diseases and the largest percentage of deaths was among those with chronic myeloid leukemia (CML).


  • An Application of Fuzzy ANOVA on Field of Agricultural Machinery

    by Zeynep Gökkuş, Sevil Şentürk, Taner Yildiz, Elçin Yeşiloğlu Cevher