http://dspace.beu.edu.tr:8080/xmlui/handle/123456789/14858 2026-04-23T06:51:41Z http://dspace.beu.edu.tr:8080/xmlui/handle/123456789/14958 Upper and lower extremity bone segmentation with Mask R-CNN AYDIN, Ayhan; ÖZCAN, Caner Most medical image processing studies use medical images to detect and measure the structure of organs and bones. The segmentation of image data is of great importance for the determination of the area to be studied and for the reduction of the size of the data to be studied. Working with image data creates an exponentially increasing workload depending on the size and number of images and requires high computing power using machine learning methods. Our study aims to achieve high success in bone segmentation, the first step in medical object detection studies. In many situations and cases, such as fractures and age estimation, the humerus and radius of the upper extremity and the femur and tibia of the lower extremity of the human skeleton provide data. In our bone segmentation study on X-RAY images, 160 images from one hundred patients were collected using data compiled from accessible databases. A segmentation result with an average accuracy of 0.981 was obtained using the Mask R-CNN method with the resnet50 architecture 2024-01-01T00:00:00Z http://dspace.beu.edu.tr:8080/xmlui/handle/123456789/14957 Explicit Formulas for Optimum Parameters of Viscoelastic-type Tuned Mass Dampers KÖSEN, Mehmet Ali; TEKİN, Gülçin Tuned mass dampers (TMDs) are passive vibration control devices that are attached to a primary system to reduce the dynamic vibrations under exciting motion. The Voigt-type TMD, which is the most widely used one, is known as a standard model of dynamic vibration absorber (DVA). The purpose of this study is to improve the vibration control performance of passive control devices by using viscoelastic-type tuned mass dampers (V-TMDs). The study adopts the Zener model to represent the viscoelastic behavior of V-TMD. In this study, the fixed-point method is used to determine the optimum parameters of a V-TMD. The displacement amplification factor (DAF) of the coupled system is obtained in the frequency domain. The optimal parameters of the V-TMD system attached to an undamped single degree-of-freedom (SDOF) main system are obtained by minimizing the DAF (symbolized with 𝛽�) under the effect of base excitation. The optimum parameters, such as damping ratio (ξ) and stiffness ratio (𝜅�) of the coupled system are derived, and explicit expressions corresponding to the optimum parameters are presented for engineering design. Moreover, the change in DAF values for different mass ratios (µ) is also discussed. It is proven that V-TMD is very effective in reducing the amplitudes of vibration. The study also provides valuable insights for engineering practitioners who want to design and implement V-TMDs for vibration control applications because accurate expressions, which are simple and easy to use, are derived in order to obtain optimum parameters, and step-by-step procedures are explained 2024-01-01T00:00:00Z http://dspace.beu.edu.tr:8080/xmlui/handle/123456789/14956 Using Machine Learning Algorithms to Analyze Customer Churn with Commissions Rate for Stocks in Brokerage Firms and Banks KAYA, Hakan Stock commission rates of banks and brokerage firms are a critical factor for investors. These rates affect the cost of stock investments. It's crucial to highlight the significance of stock commission rates in brokerage firms and banks, as well as the factors that influence their determination. This article aims to draw attention to the study's focus on customer churn and commission rates within the financial industry. Previous research has mainly focused on identifying the key variables affecting customer churn without considering its impact on forecast accuracy. This work has two primary research goals: first one is to investigate how commission rates affect the accuracy of customer churn prediction in brokerage firms and the banking sector using machine learning models, and second one is to compare and evaluate the most effective machine learning approaches for predicting customer churn. The customer churn management approach was enhanced through the analysis of a data set obtained from a bank and brokerage firm. This data set, comprised of 7816 entries and 14 columns, reflects the firm's transactions and was sourced from a publicly accessible database. The analysis employed Decision Tree, Random Forest, K-NN, Gaussian NB, and XGBoost algorithms to evaluate performance using three accuracy measures. Two approaches are included for model creation. According to the first analysis results, the Gaussian NB, for second approach the K-NN algorithms gave the best result. 2024-01-01T00:00:00Z http://dspace.beu.edu.tr:8080/xmlui/handle/123456789/14955 Gamma-Ray Shielding Capacity of Bi2O3-SiO2-B2O3 Glass Powders with Different Bi2O3 Contents SENGUL, Aycan; AKKURT, İskender The study investigated the shielding properties of Bi2O3-SiO2-B2O3 glass powders with varying Bi2O3 levels (45-60 mass%) against ionizing radiation using GAMOS (version 6.2). The simulation geometry produced by GAMOS was validated by comparing the results to conventional XCOM data for mass attenuation coefficients of glass particles. The Monte Carlo simulations were used to score photons that traveled in an absorber within the energy range of 0.01 MeV to 20 MeV, depending on the parameter under study. The simulation model involved a monoenergetic point source producing a pencil beam, absorber, and detector. We have calculated the mass attenuation coefficient (MAC), Half-value layer (HVL), Tenth-value layer (TVL), and Mean Free Path (MFP). The greatest linear attenuation coefficients in the whole energy range are related to 60Bi and the lowest were to 45Bi. The obtained results were compared, and these results are in good agreement with the obtained values from the XCOM program. 2024-01-01T00:00:00Z