MEF GCRIS Premium Database
The scientific memory of MEF University. Publications, projects, and researchers—all in one place. The heart of open science beats here. 'Open Science. Visible Impact.'
Most Viewed Publications
Most Downloaded Publications
Recent Submitted Publications
Citation - Scopus: 2
Turcoins: Turkish Republic Coin Dataset
(IEEE, 2021) Gökberk, Berk; Akarun, Lale; Temiz, Hüseyin
In this paper, we present a novel and comprehensive dataset which contains Turkish Republic coins minted since 1924 and present a deep learning based system that can automatically classify coins. The proposed dataset consists of 11080 coin images from 138 different classes. To classify coins, we utilize a pre-trained neural network (ResNet50) which is pre-trained on ImageNet. We train the pre-trained neural networks on our dataset by transfer learning. The imbalanced nature of the dataset causes the classifier to show lower performance in classes with fewer samples. To alleviate the imbalance problem, we propose a StyleGAN2-based augmentation method providing realisticfake coins for rare classes. The dataset will be published in http://turcoins.
Citation - WoS: 22
Citation - Scopus: 25
Perceptions of Dating Violence: Assessment and Antecedents
(SAGE Publications, 2020) Toplu-Demirtaş, Ezgi; Fincham, Frank D.; Öztemür, Gizem
Challenging perceptions of violence is crucial to prevent dating violence (DV), because such perceptions intervene in the organization and interpretation of violent events. However, these perceptions have received limited attention. This likely reflects the lack of a psychometric tool to do so. The current study had two purposes: to develop a measure of perceptions of psychological, sexual, and physical DV, and to explore how vertical collectivism, through hostile sexism and violence myth acceptance, shapes perceptions of DV. A total of 491 college students (55.3% women; M = 20.76 years, SD = 1.77 years) completed measures of the vertical collectivism, hostile sexism, domestic violence myth acceptance, and perceptions of DV. The results of exploratory factor analyses revealed a 15-item single-factor measure of perceptions of DV as initial construct validity, which had satisfactory internal consistency. A gender difference emerged in perceptions of DV; college women perceived psychological, sexual, and physical DV as more serious compared with college men. Moreover, the association between vertical collectivism and perceptions of DV was serially mediated via hostile sexism and violence myth acceptance. The findings are discussed in terms of previous research and the need to address the role of vertical collectivism in sexism, myth acceptance, and perceptions of violence in prevention/intervention efforts to reduce vulnerability to DV perpetration and victimization. Several recommendations are outlined to facilitate future research.
The Architecture of a Library in a Digital World
(Emerald Group Publishing Limited, 2016) Çimen, Ertuğrul
Introduction: As a librarian with over 25 years of experience, I can say that the past five years have seen more rapid change than the rest of my time in the field. Technological disruptions are changing how libraries are used and how they provide resources to students. These rapid changes have brought opportunities and transformation as well as challenges and uncertainty. University libraries are at the forefront of these changes and, I believe, have a responsibility to assist in developing pathways for both university and public libraries to negotiate through the uncertainty that currently exists. What I aim to put in place at MEF University, and then use as a model for other libraries converting to the needs of digital learners is as follows. My vision is to produce innovative and “smart” high-quality information services in adherence to academic ethical rules; to undertake a leadership role among national and regional academic libraries through individual and institutional cooperation and effective use of communication channels. The Architecture of a Library in a Digital World : With regard to architecture, how libraries are set up has gone through a radical change. At MEF, in order to accommodate the digital needs of today’s students, the library has been specifically designed to incorporate workspaces with access to electric points and a strong Wi-Fi connection. The Successes of Using Digital Materials: While some hard copy books and journals are available in the MEF Library, the majority of resources are electronic. Students and instructors are provided with online access to digital resources, which they can access 24 hours a day, seven days a week. The Challenges of Using Digital Materials: While we have seen successes in our architectural model and also in our digital access model, challenges are also arising. What is emerging is that neither publishers nor institutional consumers seem to be quite ready for the shifting needs that have taken place due to the digital revolution. Academic Integrity in a Digital World: Finally, it is important to touch upon issues of academic integrity. More than ever, it is easier to fall into plagiarism when using digital materials, whether intentionally or unintentionally. Students need clear training on understanding what academic integrity encompasses, and how to avoid plagiarism. Conclusion : What libraries look like, how they are used, and how they are stocked and lend books has changed rapidly over the past twenty years. This will continue to change at an exponential rate as new technological modes of accessing knowledge and learning emerge.
Determination of Alzheimer's Disease Stages by Artificial Learning Algorithms
(Lifescience Global, 2025) Bulut, Nurgül; Çakar, Tuna; Arslan, İlker; Akıncı, Zeynep Karaoğlu; Oner, Kevser Setenay
Introduction: This study aims to determine the stages of Alzheimer's disease (AD) using different machine learning algorithms, and compares the performance of these models. Methods: Demographic, genetic, and neurocognitive inventory data from the National Alzheimer's Coordinating Center (NACC) database as well as brain volume/thickness data from magnetic resonance imaging (MRI) scans were used. Deep Neural Networks, Ordinal Logistic Regression, Random Forest, Gaussian Naive Bayes, XGBoost, and LightGBM models were used to identify four different ordinal stages of AD. Results: Although the performance measures of the developed models were similar, the highest classification rate of AD stages was achieved by the Random Forest model (accuracy: 0.86; F1 score: 0.86; AUC: 0.95). The outputs of the model with the best performance were explained by the SHapley Addictive exPlanations (SHAP) method. Conclusions: This indicates that non-invasive markers and machine learning models can be used effectively in early diagnosis and decision support systems to predict stages of AD. © 2025 Elsevier B.V., All rights reserved.