Altınbaş University Institutional Repository
Not a member yet
5805 research outputs found
Sort by
Designing a system to recognize main arabic dialects
Get PDFIdentifying dialects can be considered as one of the recently attracted researchers. This study concentrates on recognizing between famous Arabic dialects based on speeches or talks. Three types of Arabic talks are considered here namely Arabian Peninsula Dialect (APD), Egyptian Dialect (ED), and Levantine Dialect (LD). We propose three artificial neural network models to identify between the three types of talking. These models are based on the Multi-Layer Perceptron (MLP) network, Convolutional Neural Network (CNN), and Deep Recurrent Neural Network (DRNN). Furthermore, comparisons between the three proposed models are provided. So, a comprehensive study is presented in this paper. Spoken Arabic Regional Archive (SARA) dataset is employed. It is prepared and divided into three groups. These are the Original SARA (OSARA), Filtered SARA (FSARA), and Mixed SARA (MSARA), which consists of both the OSARA and FSARA. According to the results, it has been found that the best accuracy of 90.70% is for the proposed DRNN model with the MSARA group of the employed dataset
How capital structure affects stock prices, evidence from Palestine
Get PDFThis research delves into the intricate relationship between capital structure and stock prices
within the context of Palestinian listed companies. This study explores the impact of various
financial metrics on stock prices. Through a quantitative approach, the research scrutinizes
the interplay of variables such as leverage ratios, returns on assets (ROA), earnings, bookto-market value, Dividends, asset tangibility and firm size in influencing stock prices.
The findings illuminate several critical insights. Capital structure is found to exert a negative
influence on stock prices, suggesting that changes in leverage ratios can significantly impact
share values. Conversely, earnings per share (EPS), along with the profitability and size of
firms, demonstrates a positive relationship with stock prices, indicating that strong financial
performance can drive share price growth.
The research uncovers essential implications for investors, emphasizing the importance of
accounting information, particularly leverage ratios, ROA, earnings, and company size, in
making informed investment decisions. It highlights the significance of diversifying
investment portfolios and gaining a comprehensive understanding of the intricate factors
influencing stock prices. Moreover, the study underscores the need for companies to
maintain transparency in their financial reporting, ensuring that investors can access vital
information to make well-informed decisions. While offering valuable insights into the Palestinian stock market, this research also opens
the door for future studies to explore additional variables and alternative methodologies to
further enhance our comprehension of this complex relationship
Excited state dependent fast switching NLO behavior investigation of sp2 hybridized donor crystal as D-?-A push–pull switches
Get PDFThis research focused on the investigating electronic and optical properties of designed chromophores (TPTP1-TPTP5) to involve their comprehensive analysis, including geometry optimization, UV–Vis spectroscopy, analysis of the transition density matrix (TDM), and exploration of their nonlinear optical (NLO) responses. The chromophore TPTP1 and TPTP2 exhibit significant transitions, making them suitable for optical switching applications. The chromophore TPTP5 stood out with high values for linear polarizability (, 8.53 × 10-24 esu), first order polarizability β0 (3.86 × 10-24 esu), and second order hyperpolarizability (γ0, 6.41 × 10-24 esu), making it notable for its nonlinear optical response. A positive correlation was observed between their vertical ionization potential (VIP) and the γ0 related NLO response, indicating that higher VIP values correspond to stronger γ0 responses. Their UV–Vis spectroscopy was employed to examine the absorption properties of the chromophores, revealing the wavelengths (λmax) at which they absorbed light and their potential for light harvesting applications. The analysis of the TDM allowed for a deeper understanding of the redistribution of electron density during electronic transitions within the chromophores. This analysis provided valuable insights into the characteristics and nature of their excited states. Additionally, the research investigated the NLO responses of the chromophores, particularly focusing on their third harmonic generation (THG) properties. These NLO properties are crucial for potential applications in optical switches, frequency conversion, and optical signal processing. Overall, the findings from this research contribute to a comprehensive understanding of the electronic and optical properties of the designed chromophores. The obtained results open up new possibilities for their utilization in various technological fields, including light harvesting, photonics, and nonlinear optics
Artık Ağ Tabanlı Uygulamayla Gözlerde Bulunan Bakterilerin Sınıflandırılması
No full textAraştırmada, ResNet mimarisi kullanılarak TensorFlow ve Keras kütüphaneleri kullanılarak bir derin öğrenme modeli oluşturulmuştur. Çalışmada 6 farklı bakteri sınıfı için toplamda 689 adet bakteri resmi veri kümesi olarak kullanılmıştır. Yazılım tasarımı, veri ön işleme, model oluşturma ve eğitim adımlarını içermektedir. Veri ön işleme aşamasında, resimler normalize edilmiş ve boyutlandırılmıştır. Model oluşturma aşamasında, ResNet mimarisi tercih edilmiştir çünkü derin ağların daha iyi öğrenme yetenekleri sunabileceği bilinmektedir. Model eğitimi sırasında, eğitim verisi üzerinde iteratif bir yaklaşım benimsenmiş ve optimize edici işlevler kullanılarak ağın ağırlıkları ayarlanmıştır. Sonuçlar, tasarlanan yazılımın %83,33 doğruluk oranı ile bakteri resimlerini başarılı bir şekilde sınıflandırdığını göstermektedir. Bu sonuçlar, derin öğrenme tekniklerinin biyomedikal görüntü analizinde potansiyelini vurgulamaktadır. Bu çalışma, bakteri sınıflandırma konusunda daha geniş veri kümeleri ve daha gelişmiş özellik mühendisliği tekniklerinin entegrasyonunu içerecek şekilde genişletilebilir
Challenging Vavricka: questioning compatibility of the mandatory tetanus vaccination with ECHR
No full textThe compatibility of mandatory vaccinations with human rights has become a very current issue with the COVID-19 pandemic and the Vavřička ruling by the European Court of Human Rights. This ruling has faced criticism for not conducting examinations related to disease and vaccines based on direct scientific evidence. In this analysis, an assessment will be made based on direct scientific evidence about tetanus and its vaccine.
The prevailing reason for mandatory tetanus vaccination is to protect the health of the vaccinated individual. Competent adults have the right to refuse treatment. This rule also applies to preventive medical interventions, including tetanus vaccination. As a rule, parents are entitled to give consent for medical interventions on their children. If an immediate and serious threat permanently endangers the minor's life, medical intervention can be carried out against the parents' will. The limitation of parental autonomy is more disputed when the minor's life is not immediately threatened. With respect to tetanus vaccination as a preventive medical intervention, it does not eliminate an immediate and serious risk of harm. As a result, interference with the parent's discretion on tetanus vaccination as a preventive medical intervention should be evaluated for its compatibility with the current legal approach to medical interventions on minors and patient rights
Design and implementation of an autonomous vehicle enhanced by advanced driver assistance systems (ADAS) using ML
No full textThis thesis discusses the design and implementation of an autonomous vehicle enhanced
with advance driver assistance systems (ADAS) using machine learning. This vehicle can
be classified as an educational platform suitable for researchers and specialists in the field
of autonomous vehicles. Its structure can be easily modified to meet the needs of researchers,
and it can be reprogrammed with ease. The work details the construction of the vehicle,
including the chassis structure, suspension system, steering system, brakes, and the anti-lock
braking system (ABS). Control of the vehicle is achieved through a mobile phone using a
control program developed with MIT App Inventor, allowing wireless Bluetooth
communication for driving. The thesis also covers the vehicle's key tasks, such as path
planning and navigation using a specialized algorithm for selecting the shortest path to the
destination. The vehicle is equipped with ultrasonic sensors distributed around it to detect
both stationary and moving obstacles. Additionally, a LIDAR sensor is used for obstacle
detection. A machine learning model is created to sense obstacles, trained on data collected
from various sensors and scenario, and used to implement autonomous driving in simulation
and augmented reality. The results demonstrate the vehicle's ability to navigate obstacles
during its journey. Finally, the vehicle can recognize different traffic signs, trained using
machine learning on a dataset of over 50,000 samples of 43 classes of German traffic signs. The model is tested for visualization and through the vehicle's camera, enabling it to
recognize and respond to all traffic signs appropriately.Bu tez, makine öğrenimini kullanan gelişmiş sürücü destek sistemleri (ADAS) ile
geliştirilmiş otonom bir aracın tasarımını ve uygulamasını tartışmaktadır. Bu araç, otonom
araçlar alanında çalışan araştırmacılara ve uzmanlara uygun bir eğitim platformu olarak
sınıflandırılabilir. Yapısı araştırmacıların ihtiyaçlarını karşılayacak şekilde kolaylıkla
değiştirilebilir ve kolaylıkla yeniden programlanabilir. Çalışmada şasi yapısı, süspansiyon
sistemi, direksiyon sistemi, frenler ve kilitlenmeyi önleyici fren sistemi (ABS) dahil olmak
üzere aracın yapısı ayrıntılarıyla anlatılıyor. Aracın kontrolü, MIT App Inventor ile
geliştirilen ve sürüş için kablosuz Bluetooth iletişimine olanak tanıyan bir kontrol programı
kullanılarak cep telefonu aracılığıyla sağlanıyor. Tez ayrıca, hedefe giden en kısa yolu
seçmek için özel bir algoritma kullanarak yol planlama ve navigasyon gibi aracın temel
görevlerini de kapsamaktadır. Araç, hem sabit hem de hareketli engelleri tespit etmek için
etrafına dağıtılmış ultrasonik sensörlerle donatılmıştır. Ayrıca engel tespiti için LIDAR
sensörü kullanılıyor. Engelleri algılamak için bir makine öğrenimi modeli oluşturuluyor,
çeşitli sensörlerden ve senaryolardan toplanan verilerle eğitiliyor ve simülasyonda ve
artırılmış gerçeklikte otonom sürüşü uygulamak için kullanılıyor. Sonuçlar, aracın yolculuğu
sırasında engelleri aşma yeteneğini gösteriyor. Son olarak araç, 43 sınıf Alman trafik
işaretinden oluşan 50.000'den fazla örnekten oluşan bir veri kümesi üzerinde makine
öğrenimi kullanılarak eğitilen farklı trafik işaretlerini tanıyabiliyor. Model, görselleştirme açısından ve aracın kamerası aracılığıyla test edilerek tüm trafik işaretlerini uygun şekilde
tanıması ve bunlara yanıt vermesi sağlandı
Numerical analysis for a computer immersion-cooling system
No full textThis study presents a unique forced flow and heat sink cooling system and technique for a single-phase immersed cooling system. Computer systems can use electricity more effectively if electronics cooling systems are done more appropriately. Computers and other electronic devices could be cooled via immersion cooling by immersing them in a thermally conductive liquid or coolant. The CPU surface is covered with a straight -fin heat sink, and the mainboard is immersed in NOVESC 3M 649, a designed fluid that can disperse heat while utilizing immersed cooling. The simulation software program results show that increasing the number of fins from 5 to 9 led to an increase in temperature, velocity, and pressure. However, the highest dissipated power was obtained when using eight fins; therefore, increasing the number of fins is considered ineffective due to the increasing temperature
An improved deep CNN for early breast cancer detection
No full textVolume editors Rasheed J., Abu-Mahfouz A.M., Abu-Mahfouz A.M., Fahim M.Over the past several decades, breast cancer has emerged as one of the most devastating illnesses globally. Globally, breast cancer is the second highest cause of mortality among all forms of cancer. Early detection, which permits the total elimination of cancer by surgery or treatment, is one of the most efficient techniques for treating cancer. Thermography, ultrasonography, and mammography are among the different technologies created for the goal of breast cancer screening. Utilizing image processing and deep learning methods, this technology can boost the radiologist’s ability to effectively detect chest anomalies. This study advises upgrading the breast cancer detection approach using a Deep Convolutional Neural Network (DCNN) to offer precise and quick findings. Furthermore, this study separates itself from the previous one by adopting a DCNN with 12 stacked processing layers. The implementation of a 12-layered Convolutional Neural Network (CNN) considerably enhanced the precision of breast cancer diagnosis and detection. We applied the Mini Mammographic Database (MIAS) to examine the efficacy of the suggested technique. Nevertheless, the acquired data reveals that Deep CNN attained a spectacular accuracy rate of 99.1%, resulting in excellent consequences. In addition, the DPD-DCNN achieved the greatest degree of accuracy when compared to similar trials
Evaluation of the clinical performance of different occlusal device materials
No full textStatement of problem: Computer aided technologies have been used to fabricate occlusal devices. However, the clinical behavior of the newly developed materials developed for occlusal devices is unknown.
Purpose: The purpose of this prospective, double-blind study was to assess the clinical efficacy of recently introduced computer-aided design and computer-aided manufacturing (CAD-CAM) materials for the fabrication of occlusal devices.
Material and methods: A total of 24 participants were divided randomly into 2 study groups; polyetheretherketone (PEEK) and polymethyl methacrylate (PMMA), and a control group (CG). Conventional impressions and gypsum casts were obtained from all participants. In the study groups, the casts were digitalized with an extraoral digital scanner, designed with a software program (Bite Occlusal Device Module; exocad GmbH) and milled from PEEK and PMMA blocks. Clear resin sheets were used for occlusal device fabrication in the CG. The baseline measurements were made during the initial appointments. After 6 months, the participants returned for follow-up evaluations. Clinical performance based on surface roughness, wear of the antagonist teeth, occlusal device fit and therapeutic effect, as well as participant satisfaction were compared using the 1-way ANOVA test between the main groups (α=.05). The post hoc and Kruskal Wallis-H tests were used to compare the nonparametric group.
Results: The therapeutic effects of the occlusal devices did not differ. All participants showed improvement in palpation and mandibular movement scores, but no statistically significant differences were found among the groups (P>.05). PEEK and PMMA had statistically less surface wear than CG (P.05). The control group had the best fit (P<.001).
Conclusions: Recent CAD-CAM materials exhibit clinically acceptable outcomes, and their performance is comparable with that of traditional materials. CAD-CAM materials appear suitable in terms of accuracy, surface wear, and therapeutic efficacy.Funding Agency: İstanbul University.
Grant Number: 35910
Malignancies and lymphoproliferations in children with primary immune deficiency-a single-center experience
No full textPrimary immune deficiencies (PIDs) are rare genetic disorders characterized by impaired immune function, leading to frequent infections and immune dysregulation. Studies have shown that individuals with PID are at an increased risk of developing malignancies and lymphoproliferative disorders compared with the general population. In this single-center study, we aimed to analyze the occurrence of malignancies and lymphoproliferations in children diagnosed with PID. We retrospectively analyzed the medical records of 550 pediatric patients diagnosed with PIDs at our center. Among them, 17 (3,0%) patients were identified with malignancy and/or benign lymphoproliferation. Eight of the 17 patients (47.0%) had immune dysregulatory diseases, whereas ataxia-telangiectasia was the second most common PID associated with malignancy and/or benign lymphoproliferation (n = 5, 29.4%). Lymphoma was the predominant malignancy (n = 11, 64.7%), and Epstein-Barr virus was identified as the most common viral agent associated with malignancy and/or benign lymphoproliferation in patients with PID (n = 8, 47.0%). Our study highlights the association between PID and malignancies/lymphoproliferations, with immune dysregulation syndromes being the most common subclass associated with malignancies/lymphoproliferations. Early diagnosis, multidisciplinary management, and regular surveillance are crucial in improving patient outcomes and saving lives