Istanbul Technical University
Ulusal Üniversitelerarası Açık Erişim Sistemi - İstanbul Teknik ÜniversitesiNot a member yet
67356 research outputs found
Sort by
Adaptive Robust L2 Loss Function Using Fractional Calculus
No full texthttps://doi.org/10.1007/978-3-031-83191-1_1
Prioritization of Human-Centric and Sustainable City Criteria by Proportional Spherical Fuzzy Analytic Hierarchy Process
No full textHuman-centric and sustainable cities have become one of the most popular research areas today. They are based on a holistic approach that aims to both improve people’s quality of life and ensure environmental sustainability. The weighting of numerous tangible and intangible criteria used in the performance evaluation of these cities has emerged as an important problem. The symmetrical representation of vague and imprecise data using fuzzy set theory is an absolute necessity for a successful weighting process. In this study, symmetrical representation of especially intangible criteria is carried out with proportional spherical fuzzy sets. Proportional spherical fuzzy sets provide significant convenience to the expert in determining the membership, non-membership and hesitancy degrees and ensure that the assigned values are more accurate and consistent. In the study, five main human centric and sustainable city (HCSC) criteria and 26 HCSC sub-criteria determined from the literature were weighted. In addition, criteria weights were obtained by classical spherical fuzzy analytic hierarchy process (AHP) for comparison purpose. The first four most important HCSC criteria were determined as water management and conservation, employment/unemployment rate, carbon emission reduction strategies and economic sustainability, respectively.https://doi.org/10.3390/sym1702020
Comparative Assessment of CNN and Transformer U‐Nets in Multiple Sclerosis Lesion Segmentation
No full textABSTRACTMultiple sclerosis (MS) is a chronic autoimmune disease that causes lesions in the central nervous system. Accurate segmentation and quantification of these lesions are essential to monitor disease progression and evaluate treatments. Several architectures are used for such studies, the most popular being U‐Net‐based models. Therefore, this study compares CNN‐based and Transformer‐based U‐Net architectures for MS lesion segmentation. Six U‐Net architectures based on CNN and transformer, namely U‐Net, R2U‐Net, V‐Net, Attention U‐Net, TransUNet, and SwinUNet, were trained and evaluated on two MS datasets, ISBI2015 and MSSEG2016. T1‐w, T2‐w, and FLAIR sequences were jointly used to obtain more detailed features. A hybrid loss function, which involves the addition of focal Tversky and Dice losses, was exploited to improve the performance of models. This study was carried out in three steps. First, each model was trained separately and evaluated in each dataset. Second, each model was trained on the ISBI2015 dataset and evaluated on the MSSEG2016 dataset and vice versa. Finally, these two datasets were combined to increase the training samples and assessed on the ISBI2015 dataset. Accordingly, the R2U‐Net and the V‐Net models (CNN‐based) achieved the best ISBI scores among the other models. The R2U‐Net model achieved the best ISBI scores in the first and last steps with average scores of 92.82 and 92.91, while the V‐Net model achieved the best ISBI score in the second step with an average score of 91.28. Our results show that CNN‐based models surpass the Transformer‐based U‐Net models in most metrics for MS lesion segmentation.https://doi.org/10.1002/ima.7014
Birational Equivalences and Generalized Weyl Algebra
No full textWe calculate suitably localized Hochschild homologies of various quantum groups and Podleś spheres after realizing them as generalized Weyl algebras (GWAs). We use the fact that every GWA is birationally equivalent to a smash product with a 1-torus. We also address and solve the birational equivalence problem, and the birational smoothness problem for GWAs.https://doi.org/10.3842/sigma.2025.063https://dx.doi.org/10.48550/arxiv.2009.14801http://arxiv.org/abs/2009.14801https://zbmath.org/808326
RSSI Fingerprint-Based Indoor Localization Solutions Using Machine Learning Algorithms: A Comprehensive Review
No full textWith the development of technologies and the growing need for accurate positioning inside buildings, the localization method based on Received Signal Strength Indicator (RSSI) fingerprinting is becoming increasingly popular. Its popularity is explained by the relative simplicity of implementation, low cost and the ability to use existing wireless infrastructure. This review article covers all the key aspects of building such systems: from the wireless communication technology and the creation of a radiomap to data preprocessing methods and model training using machine learning (ML) and deep learning (DL) algorithms. Specific recommendations are provided for each stage that can be useful for both researchers and practicing engineers. Particular attention is paid to such important issues as RSSI signal instability, the impact of multipath propagation, differences between devices and system scalability issues. In conclusion, the review highlights the most promising areas for further research. For smart cities, the approaches and recommendations presented in the review contribute to the development of urban services by combining indoor positioning systems with IoT platforms for automation, transport and energy management.https://doi.org/10.3390/smartcities805015
Targeting Ubiquitin-Specific Protease 7 (USP7): A Pharmacophore-Guided Drug Repurposing and Physics-Based Molecular Simulation Study
No full textAbstractUbiquitin-Specific Protease 7 (USP7) has emerged as a critical therapeutic target in cancer due to its regulatory roles of tumor suppressors, oncoproteins, and epigenetic modifiers. In this study, we present a comprehensive drug repurposing strategy to identify potential USP7 inhibitors from a curated library of 6,654 FDA-approved and investigational small molecules. Using structure-based pharmacophore models derived from USP7-inhibitor crystal structures, we conducted virtual screening to select candidates with favorable pharmacophoric alignment. The top 100 hits were subjected to short 10 ns molecular dynamics (MD) simulations and MM/GBSA binding energy calculations, narrowing down to 36 promising ligands. These were further evaluated through longer (100 ns) MD simulations, binding energy refinement, ligand clustering based on molecular fingerprints, and cancer-specific activity predictions using a binary QSAR model. Twelve compounds demonstrated favorable binding profiles, structural diversity, and high predicted anticancer activity, such as xantifibrate, etofylline clofibrate, and carafiban. This multi-tiered hybrid virtual screening pipeline highlights the potential of drug repurposing for the rapid identification of USP7 inhibitors, offering a cost-effective path toward novel anticancer therapeutics.https://doi.org/10.1021/acsomega.5c03150https://doi.org/10.1101/2025.04.07.647568https://pmc.ncbi.nlm.nih.gov/articles/PMC12593964/https://pubmed.ncbi.nlm.nih.gov/41210757
The evolution and impact of Karagöz plays’ sound in early Republican Türkiye: a focus on Hayali Küçük Ali
No full textThis article examines the process through which Karagöz plays became recognized as cultural heritage in the early years of the Republic of Türkiye, arguing that Karagöz should be regarded not only as a visual but also as an auditory form of heritage. By analyzing the role of auditory components in cultural continuity and transformation, the study aims to reconsider definitions of heritage through the lens of sound. It investigates how state policies, technological developments, and ideological shifts reshaped Karagöz performances in the transformation of their musical and sonic elements into cultural heritage, and how these changes influenced the case of the puppeteer Hayali Küçük Ali. As part of the research, two musical pieces selected from the Karagöz repertoire and Küçük Ali’s performances of Aşıklık and Salıncak plays were taken as case studies; transcriptions of archival recordings from Turkish Radio and Television (TRT) were compared with Cevdet Kudret’s written texts. The analysis was conducted within the framework of parameters such as character voicing, musical elements, auditory environment, improvisation, comedic devices, ideological reflections, and contextual relations. The findings reveal that sound functions both as a medium of transmission and reproduction, and that cultural memory is shaped within these auditory sites of remembrance. The modernization of Karagöz unfolded through a sound-centered logic of adaptation, evolving into a multilayered auditory heritage extending from radio broadcasts to digital platforms. Thus, the plays have undergone a sound-oriented reinterpretation in terms of both performance and preservation. In conclusion, this study foregrounds the auditory and musical dimensions of Karagöz to examine the relationship between performance memory, ideological transformation, and technological mediation, inviting a reconsideration of the notion of an “auditory Karagöz heritage” in relation to cultural transmission and representation.https://doi.org/10.12975/rastmd.2025133
Multifunctional Ti3C2Tx MXene/carbon nanotube interlayer as a polysulfide electrocatalyst with a high sulfur loading cathode in pre-lithiation Si/S batteries
No full texta) Self discharge capacity b) open circuit voltage profiles showing self-discharge behavior for pristine, CNT and Ti3C2Tx/CNT separators.https://doi.org/10.1039/d5cy00674
Central Bank Digital Currencies (CBDCs): a countermeasure to Anti-Money Laundering (AML) challenges posed by cryptocurrencies?
No full texthttps://doi.org/10.1007/s42521-025-00132-
A green approach to prepare graphene-coated flexible and conductive cotton fabrics for motion sensing
No full textAbstract Electrically conductive textiles have revolutionized wearable electronics, integrating them to create interactive and intelligent clothing. Graphene is an ideal candidate for increasing the electrical properties of textiles owing to its mechanical flexibility and high electrical conductivity. In this work, electrically conductive cotton fabrics were prepared through the dip-dry technique using a solution containing graphene nanoplatelets (GNPs), sodium dodecyl benzene sulfonate (SDBS), and polyvinyl alcohol (PVA). The sheet resistance of seven times coated cotton fabrics was ~ 1.9 (kΩ/sq), which showed their sensing capabilities. In addition, the flexibility and durability of the coated cotton fabrics were evaluated to certify their suitability for wearable applications using a cyclic test up to around 2300. Prepared conductive cotton fabrics were tested in four main body joints with different angles (45°, 90° and 120°), i.e. wrist, elbow, knee, and finger. The coated fabrics can detect mechanical actions such as bending/unbending, and stretching/relaxation as skin-mounted strain sensors.https://doi.org/10.1007/s10853-025-10592-