Hacettepe Üniversitesi Açık Erişim Sistemi (HÜAES)

  • HÜAES, Hacettepe Üniversitesi bünyesinde üretilen kitap, makale, tez, bildiri, rapor gibi tüm akademik kaynakları uluslararası standartlarda dijital ortamda depolar, etkisini artırmak için telif haklarına uygun olarak Açık Erişime sunar.
 

Communities in DSpace

Select a community to browse its collections.

Now showing 1 - 5 of 22

Recent Submissions

Thumbnail Image
Item
Data-Driven Software Engineering: The Integral Role Of Generative Ai In Transforming Software Development Processes
(Fen Bilimleri Enstitüsü, 2025) Yalçıner, Aybüke; Bilgisayar Mühendisliği
This thesis examines the transformative convergence of data science and software engineering, with a particular emphasis on how generative AI technologies are revolutionizing traditional software development processes. The thesis introduces a framework for the ethical and effective implementation of generative AI across the software development lifecycle. There has been a significant impact of emerging technologies on software engineering processes over the past few years. Consequently, there has been a shift from experience based to data-driven decision-making. This transition towards data-driven decision making has led to more precise and reliable decisions, resulting in more effective and efficient software engineering processes with reduced rework. Our study consists of two stages. The first stage includes a comprehensive systematic literature review on the applications of data-driven methodologies in software engineering processes from the past decade. Through the analysis of 34 primary studies, we identified the domains where the use of data-driven methodologies is most prevalent. The analysis showed that data driven approaches are widely applied in software management and software testing within software engineering processes. Researchers have been utilizing various subfields of artificial intelligence, including machine learning and deep learning, to develop decision models for software engineering processes. In the second section of our study, to satisfy a gap determined as a result of the SLR, we proposed a framework for utilizing generative AI in software engineering processes to reduce costs, shorten project timelines, improve workflow efficiency, and enhance software quality. The framework outlines how data should be processed and how generative AI can be leveraged in software engineering processes, as outlined in SWEBOK v4, and presents the benefits and limitations of its application.
Thumbnail Image
Item
Thermal Modelling Of Laser Cutting Process Of Ultra High Strength Steels And Optimum Cutting Parameters Selection
(Fen Bilimleri Enstitüsü, 2025) Çetin, Doğukan; Makine Mühendisliği
Weapon and ammunition technologies are developing. This makes armored vehicles to be faster and more agile. To achieve agility, armored vehicles need to be lighter. Certain ballistic levels can be obtained in lower thickness, lower hull weight, and values with the help of Ultra High Strength steels. These steels are manufactured via quenching and tempering operations. Laser cutting, plasma cutting, waterjet cutting, bending and machining are the most common ways of manufacturing with ultra high strength steels. During thermal cutting operations, the cutting area reaches to melting temperature. Inside of the heat affected zone it is expected to see rapidly cooling regions. In the scope of this study laser cutting process is modelled via a volumetric cylindrical heat source. Simulations are repeated for 1900W, 1500W and 1200 W laser power values and 1950 [mm/min], 1200 [mm/min] and 800 [mm/min] feed rates. Results are experimentally validated. Temperature distribution and cooling behavior are investigated via continuous cooling transformation diagrams of Protection 600T steel. As a result, 1200 W and 1200 [mm/min] coupling is selected as the optimum cutting parameter.
Thumbnail Image
Item
Çizge Boyama Problemleri Üzerine Kapsamlı Bir Araştırma
(Fen Bilimleri Enstitüsü, 2025) Ünal, Murat Erşen; Matematik
Graph theory is one of the most significant research areas in mathematics due to its wide range of practical applications. In this thesis, we focus specifically on the graph coloring problem. First, we introduce some fundamental concepts. Subsequently, we present key theoretical results concerning existing lower and upper bounds related to graph coloring. Additionally, we discuss essential concepts and foundations of computational complexity theory. Using these, we formally and informally introduce the classes of P and NP prob lems. Furthermore, we define the class of NP-complete problems and demonstrate, via the Cook-Levin Theorem, that this class is non-empty. Consequently, we establish that determin ing whether a given graph can be properly colored with a given number of colors k (where k is a natural number) is an NP-complete problem. Although noknownpolynomial-timealgorithmsolves this problem optimally, several heuris tic algorithms exist in the literature for graph coloring. Among these, we examine the Greedy, DSatur, HEA, and TabuCol algorithms in detail and provide an original comparative analysis based on randomly generated graphs. The graph coloring problem also has real-world applications. We discuss various practical implementations, with a particular focus on examination scheduling, one of its most prominent and widely used applications. In this context, we develop an original design and implement a software solution in Python using Integer Linear Programming (ILP). In conclusion, this thesis serves as a comprehensive Turkish-language resource on graph coloring and computational complexity. Moreover, the developed software will be made available for use by the Department of Mathematics, Faculty of Science, Hacettepe Univer sity.
Thumbnail Image
Item
Zoledronik Asit Modifiye Stronsiyum/Hidroksiapatit Nanopartikül İçeren Hidrojellerin Geliştirilmesi
(Fen Bilimleri Enstitüsü, 2025) Özgen, Alkin; Biyomühendislik
Within the scope of this doctoral thesis, new types of nanocomposite hydrogels consisting of zoledronic acid/strontium hydroxyapatite nanoparticles and carboxymethyl chitosan/oxidised dextran (CMC/OD) hydrogels were studied. Pure hydroxyapatite, 5%, 10% and 15% (w/w) strontium substituted strontium hydroxyapatite nanoparticles were produced and modified with zoledronic acid at ratios of 5% to 7,5% (w/w). The modified strontium hydroxyapatite nanoparticles were incorporated into CMC/OD hydrogels. Zoledronic acid modified strontium hydroxyapatite nanoparticles were characterised by Scanning Electron Microscopy (SEM), Energy Dispersive X-Ray Spectroscopy (EDS) and X-Ray Diffraction (XRD). The chemical properties of CMC/OD hydrogels were investigated by Fourier Transform Infrared Analysis (FTIR) and SEM analysis. The physical properties of CMC/OD hydrogels were determined by degradation behaviour and rheological measurements. Cell-material interactions were investigated in vitro. The results showed that the incorporation of hydroxyapatite nanoparticles into CMC/OD hydrogels would significantly improve the rheological properties. The addition of strontium to hydroxyapatite nanoparticles significantly enhanced cell proliferation. A significant increase in alkaline phosphatase (ALP) and calcium accumulation was observed with zoledronic acid modification. In conclusion, CMC/OD nanocomposite hydrogels containing zoledronic acid modified strontium hydroxyapatite show potential in orthopaedic and craniofacial applications due to their superior properties such as the ability to be easily injected into targeted areas, a strong antibacterial activity that prevents infections, and self healing capabilities that promote tissue regeneration and repair.
Thumbnail Image
Item
Fototermal, Fotodinamık Ve Kemodinamık Terapi İçin İnorganik Nanopartiküllerin Sentezi Ve Karakterizasyonu
(Fen Bilimleri Enstitüsü, 2025) Süngü Akdoğan, Çağıl Zeynep; Biyomühendislik
Metastasis and treatment resistance, which are major challenges in oncology, limit the effectiveness of current treatment protocols. Copper (Cu), due to its Fenton-like activity, and cerium (Ce), due to its high redox cycle and catalase- like activity, stand out as promising, cost effective, and readily available materials for nanoparticulate therapeutic approaches aimed at increasing reactive oxygen species (ROS) production to enhance damage to tumor cells by overcoming the resistance of the hypoxic tumor microenvironment. In this study, multifunctional synergistic therapeutic agents based on Cu and Ce were designed to overcome the current limitations of cancer treatments. These agents, developed as mesoporous copper (II) oxide nanorods and hollow cerium oxide nanoparticles, demonstrated a strong synergistic effect against tumor cells by integrating chemodynamic, photodynamic, photothermal, and starvation therapies. Through surface modifications of the nanorods and nanoparticles, multiple functionalities such as hydrogen peroxide production, catalase-like activity, oxygen bubble formation, and heat generation were achieved. Consequently, an environment was created that enhances oxidative stress in the tumor microenvironment, triggering cell death. Additionally, starvation therapy induced through glucose consumption disrupted the energy metabolism of tumor cells, increasing their sensitivity to treatment. Mesoporous copper (II) oxide nanorods (CuO) were synthesized via heterogeneous nucleation on polymeric nanoparticles and functionalized with a CaO₂ nanoshell coating (CuO@CaO₂). The decomposition of the nanoshell in aqueous media activated the Cu(I)/Cu(II) cycle through the generation of H₂O₂, significantly increasing glutathione (GSH) depletion. The decomposition of H₂O₂ through catalase-like activity resulted in the formation of O₂ bubbles, propelling the CuO@CaO₂ nanorods like nanomotors. In addition to catalase activity, these nanostructures exhibited peroxidase- and oxidase-like activities. The peroxidase-like activity of CuO@CaO₂ nanorods potentiated the chemodynamic effect in the tumor microenvironment by using self-generated H₂O₂ to enhance the production of toxic hydroxyl (•OH) radicals. When exposed to near-infrared (NIR) laser irradiation, the nanorods exhibited high photothermal conversion properties due to significant temperature increases. Through photothermal therapy (PTT), GSH depletion was enhanced, and •OH radical production was further optimized, improving chemodynamic function. The therapeutic potential was evaluated against T98G cells by loading chlorin e6 (Ce6) onto the nanorods. Using CuO@CaO₂@Ce6 nanorods, the synergistic combination of photodynamic therapy (PDT), PTT, and chemodynamic therapy (CDT) resulted in over 90% cell death in vitro. To further increase oxygen levels in the tumor microenvironment, hollow and mesoporous CeO₂ nanoparticles (H-CeO₂) exhibiting catalase-like activity were synthesized using a stage shape-templating protocol. These nanoparticles were loaded with Ce6 through adsorption and coated with a thin polydopamine (PDA) layer. The PDA shell facilitated PTT conversion upon exposure to an 808 nm NIR laser. By immobilizing glucose oxidase (GOx) onto H CeO₂@Ce6@PDA nanoparticles, glucose in the tumor microenvironment was converted into H₂O₂ and gluconic acid. The conversion of glucose into the tumor-toxic •OH radical demonstrated the CDT effect of these nanoparticles. Glucose consumption induced starvation in tumor cells, increasing ROS production in the tumor microenvironment and enhancing PDT efficacy. The in vitro synergistic effects of starvation therapy (ST), PDT, and PTT, without the use of any drugs, were tested on T98G glioblastoma cells, resulting in over 90% cell death. In conclusion, two novel synergistic therapeutic agents were developed within the scope of this thesis, representing a significant step toward designing effective cancer treatment strategies.