Welcome to IKU Academic Digital Archive System
OpenAccess@IKU is Istanbul Kultur University's Academic Digital Archive System, established in June 2014 to digitally store and provide open access to academic and artistic outputs in line with international standards and intellectual property rights. The system includes various outputs such as articles, presentations, theses, books, book chapters, reports, encyclopedias, and works of art produced by the university's faculty members and students.
IKU University is a leading institution using UniSpace.
Recent Submissions
Deep Ensemble Learning and Explainable AI for Multi-Class Classification of Earthstar Fungal Species
(MDPI, 2025) Kumru, Eda; KORKMAZ, ARAS FAHRETTİN; Ekinci, Fatih; Aydoğan, Abdullah; Güzel, Mehmet Serdar
The current study presents a multi-class, image-based classification of eight morphologically similar macroscopic Earthstar fungal species (Astraeus hygrometricus, Geastrum coronatum, G. elegans, G. fimbriatum, G. quadrifidum, G. rufescens, G. triplex, and Myriostoma coliforme) using deep learning and explainable artificial intelligence (XAI) techniques. For the first time in the literature, these species are evaluated together, providing a highly challenging dataset due to significant visual overlap. Eight different convolutional neural network (CNN) and transformer-based architectures were employed, including EfficientNetV2-M, DenseNet121, MaxViT-S, DeiT, RegNetY-8GF, MobileNetV3, EfficientNet-B3, and MnasNet. The accuracy scores of these models ranged from 86.16% to 96.23%, with EfficientNet-B3 achieving the best individual performance. To enhance interpretability, Grad-CAM and Score-CAM methods were utilised to visualise the rationale behind each classification decision. A key novelty of this study is the design of two hybrid ensemble models: EfficientNet-B3 + DeiT and DenseNet121 + MaxViT-S. These ensembles further improved classification stability, reaching 93.71% and 93.08% accuracy, respectively. Based on metric-based evaluation, the EfficientNet-B3 + DeiT model delivered the most balanced performance, with 93.83% precision, 93.72% recall, 93.73% F1-score, 99.10% specificity, a log loss of 0.2292, and an MCC of 0.9282. Moreover, this modeling approach holds potential for monitoring symbiotic fungal species in agricultural ecosystems and supporting sustainable production strategies. This research contributes to the literature by introducing a novel framework that simultaneously emphasises classification accuracy and model interpretability in fungal taxonomy. The proposed method successfully classified morphologically similar puffball species with high accuracy, while explainable AI techniques revealed biologically meaningful insights. All evaluation metrics were computed exclusively on a 10% independent test set that was entirely separate from the training and validation phases. Future work will focus on expanding the dataset with samples from diverse ecological regions and testing the method under field conditions.
Vibrational Spectroscopic, Thermophysical, And Structural Properties Of Two Antarctic Howardites: EET 87503 and QUE 97001
(Elsevier Ltd., 2025) Unsalan, Ozan; Altunayar-Unsalan, Cisem; Nogueira, Bernardo A.; Kaliwoda, Melanie; FAUSTO, RUI
This study uses optical microscopy, Raman spectroscopy, differential scanning calorimetry (DSC), and thermogravimetric analysis (TGA) to examine the Antarctic howardite meteorites EET 87503 and QUE 97001. DSC results revealed troilite phase transitions in EET and QUE at 146.66 and 147.50 °C, corresponding to 0.26 % and 0.13 % troilite content, respectively. TGA indicated minor weight loss (<1 %) in both samples, with EET showing 0.399 % and QUE 0.638 % weight loss upon heating up to 1200 °C. Raman spectroscopy confirmed the presence of key minerals, including enstatite, ferrosilite, diopside, forsterite, ilmenite, and anorthite, as well as, in the case of the QUE 97001 meteorite, monticellite, a rare magnesium-end-member silicate olivine type mineral, providing insights into the complex thermal and impact histories of these howardites. The present findings appear as a contribution to a better understanding of the mineralogy and thermal evolution of this type of meteorites, linking them to potential parent bodies such as asteroid 4Vesta.
Exploring The Frontier Of Non-coding RNAs: Biogenesis, Functional Diversity, And Disease Associations Of CircRNAs And TinyRNAs
(Springer Science and Business Media B.V., 2025) ÇALIŞIR, HİLAL-NUR; AKÇAY, SILA; ENES, BAL; KILBAŞ, PELİN ÖZFİLİZ; BUSE, ÖZDEN; ŞAHİN, BURCU AYHAN
Non-coding RNAs (ncRNAs) represent a pivotal frontier in molecular biology, expanding our understanding of genetic regulation beyond the classical central dogma. Among these, circular RNAs (circRNAs) and tiny RNAs (tyRNAs) have emerged as critical regulators with diverse roles in gene expression, cellular homeostasis, and disease pathogenesis. CircRNAs, characterized by their covalently closed structures, exhibit remarkable stability and functional versatility, including acting as microRNA sponges, transcriptional regulators, and protein scaffolds. Tiny RNAs, a newly recognized class of ncRNAs, are implicated in post-transcriptional regulation through interactions with Argonaute proteins, influencing pathways from viral defense to neurodegenerative diseases. This review synthesizes recent advancements in the biogenesis, molecular mechanisms, and clinical significance of circRNAs and tyRNAs, highlighting their potential as biomarkers and therapeutic targets in cancer, metabolic disorders, and neurological conditions. By exploring their unique regulatory landscapes, this work aims to inspire further research into the ncRNA universe and its implications for human health.
YouTube Video Comments Sentiment Analysis Using Custom NLP Model
(IEEE, 2025) ERCİKAN, YAVUZ SELİM; ELMASRY, WİSAM
This paper presents a sentiment analysis platform designed to process user comments on YouTube videos. Leveraging an LSTM-based neural network trained on large-scale datasets such as Sentiment140 and a 3-million-row Twitter sentiment dataset, the platform categorizes comments into positive, negative, or neutral sentiments. It integrates modern web technologies like React.js for the front-end, Flask for the back-end, Firebase for user authentication, and the YouTube Data API for comment retrieval. Additionally, OpenAI's language models are employed to provide advanced contextual analysis, extracting key themes and emotional trends from the comments. The system visualizes sentiment distributions using dynamic charts, offering insights valuable for content creators and researchers. This work demonstrates the effective combination of deep learning and scalable web technologies to build a user-friendly sentiment analysis solution.
APAYDIN, TUĞÇE
Dr. Öğr. Üyesi