Yıldız Technical University Research Information System
Not a member yet
91324 research outputs found
Sort by
Attention-Based Decision Fusion for Breast Cancer Classification Using Ensemble Transformers Topluluk D n st r c ler Kullanarak Dikkat Tabanli Karar Birle stirme ile Meme Kanseri Siniflandirmasi
No full textBreast cancer is a significant global health issue, and early diagnosis and accurate diagnostic processes are of great importance. In this study, deep learning-based transformer models were fine-tuned using a transfer learning approach for the classification of breast cancer histopathological images, and an attention mechanism-based decision fusion method was proposed to optimize model predictions. Experiments conducted on a widely used dataset in the literature demonstrated that the highest classification performance among individual models was achieved with an accuracy rate of 92.25%. However, using the proposed attention-based fusion method, an accuracy rate of 95% was attained on the test set. Additionally, analyses performed on an independent hidden test dataset to evaluate the model's generalization capability achieved an accuracy rate of 90%, indicating that the proposed method provides an effective solution for breast cancer classification
Bulanık AHP ve Bulanık TOPSIS Yöntemleriyle Sürdürülebilir Tedarikçi Seçimi: Modüler Mobilya Sektöründe Bir Uygulama
No full textGeogebra ve 3D Yazıcı Kalem ile Geometrik Cisimlerde Boyut Kavramına Yönelik Algının ve Geometriye Karşı Tutumun Geliştirilmesi
No full textRomantik İlişkisi Olan Bireylerde Bilinçli Farkındalık, Duygusal Zeka ve Mental İyi Oluşun İlişki Doyumu ile İlişkisi
No full textForecasting of Solar Power by LSTM Derivatives: A Comparison Study
No full textThe utilization of solar energy diminishes thereliance on non-renewable fossil fuels and mitigates climate changeconsequences by lowering carbon emissions. Photovoltaic systems are among themost used technologies for transforming solar energy into useful electricalenergy. However, power generation in solar systems varies due to numeroussystem-related factors. Therefore, the estimation of photovoltaic energy isimportant in terms of management, planning, integration and sustainability. Additionally,manual solar energy forecasting requires significant human effort and expertiseto analyse and interpret the data. This process is time-consuming and open toerrors. Thus, recently, the researchers around the world have been focusing onthe application of artificial intelligence and machine learning approaches toforecast power production in the highly unpredictable renewable energy sector.In this study, the performances of the derivatives of long-short term memory(LSTM) neural networks are investigated in one hour ahead forecasting of thesolar power output for a solar farm located in Türkiye. The aim of the study isto determine the best performing LSTM derivative for this problem anddemonstrate the application of the algorithms in a specific case study, wherethe utilized data is publicly available. Thus, the methodology presented herecan be utilized by anyone, including government agencies, to forecast solarpower output of any solar farm, making this study a significant contribution tothe existing literature.</p
Evaluation of Yunus Emre Poems Designed as Digital Poetry with Web 2.0 Tools in Terms of Language and Values Education
No full textThis study aims to assess the contribution of digital poems by Yunus Emre—a prominent figure in Turkish-Islamic culture—created using Web 2.0 tools to language and values education from the perspective of teachers. Within the scope of the research, eight poems from Yunus Emre’s Divan, selected based on their appropriateness for students' age and developmental levels, were transformed into digital poems using Canva. Skill-based Turkish language questions were appended to the end of these poems, supported by activities focused on fundamental language skills, thereby converting them into unique, technology-enhanced teaching tools.The prepared digital poems were assessed by 17 Turkish language teachers working in public secondary schools in Manisa. The teachers provided insights into the effects of digital poems on basic language skills, life perspectives, values education, friend selection, cultural development, and assessment methods. Additionally, their metaphorical perceptions of the concepts of "digital poetry" and "Yunus Emre" were examined.Teachers' feedback was collected through a survey developed by the researchers and administered via Google Survey. The research data were analyzed using content analysis. The findings indicate that the seven poems, centered around themes of love and brotherhood in Yunus Emre’s Divan, contribute positively to students' fundamental language skills, perspectives on life, values education, friend selection, and cultural development. The importance of introducing such literary works to students was emphasized. Furthermore, it was concluded that the content, developed based on insights from subject matter experts and teachers, could serve as valuable instructional materials in Turkish language education
Comparison of the Removal of Synthetic Wastewater Samples Containing Basic Blue 3 Dye Using Electrochemical and Adsorption Methods
No full textWater pollution, a significant environmental issue, is growing more urgent. This study evaluated the effectiveness of adsorption and electrocoagulation methods in removing Ba-sic Blue 3 (BB3), a common dye used in the textile industry, from water. For the adsorption process, linden tree leaves—often used for health benefits in existing literature—were employed, while in the electrocoagulation (EC) method, an aluminum electrode was used. The results show that the optimal conditions for adsorption were an initial BB3 concentration of 5 mg/L, 50 mL of 0.9 g Tilia L. adsorbent, 60 min, 180 rpm, 30 °C, and pH 10, achieving a removal efficiency of 99.21%. The optimal conditions for electrocoagulation were 1 L of 15 mg/L initial BB3, a current density of 2.64 mA/cm2, 15 mL of 0.2 M KCl, a reaction time of 90 min, a stirring speed of 100 rpm, and a pH of 10, resulting in a removal efficiency of 97.98%. The results indicate that linden leaves, a natural and sustainable material, showed a slightly higher removal percentage (99.21%) in the EC method over a shorter period (60 min). Conversely, the EC method also achieved a significant removal rate (97.98%, 90 min). In summary, both methods demonstrate strong BB3 removal capabilities and could help improve wastewater treatment processes
Evaluating greenhouse gas emissions and carbon credit potential of solid waste disposal facilities: the case of Istanbul
No full textIncreasing population and urbanization across the world are rapidly increasing waste generation, which in turn contributes significantly to global greenhouse gas emissions. In the fight against climate change, technology is important for reducing waste emissions. Hence, in this study, greenhouse gas emissions and carbon credit values of municipal solid waste (MSW) disposal facilities such as sanitary landfill, waste incineration and biogas facilities in Istanbul are analyzed for the year of 2023, using real facility data based on IPCC and Gold Standard methodologies. The results obtained are comprehensively evaluated with the related literature studies. The annual non-biogenic emissions per t of waste were found to be: 0.461 tCO2e for Kemerburgaz waste incineration; 0.457 tCO2e for Seymen LFG and 0.001 tCO2e for Kemerburgaz biogas. In addition, it is determined that the waste incineration facility has the highest carbon credit amount with 1,469,676 tCO₂e. Carbon reduction of 1,257,087 tCO₂e is achieved in the sanitary landfill facility and 16,340 tCO₂e in the biogas facility. Energy generation per t of waste was also highest in the incineration facility (0.59 MWh), compared to the biogas facility (0.19 MWh) and LFG facility (0.09 MWh). As a result, this study provides contribution to the development of sustainable waste management strategies, reduction of greenhouse gas emissions and evaluation of carbon credit potentials and to be a reference source in related fields. Graphic abstract: (Figure presented.
Enhanced Embolic Signal Analysis through Ensemble Deep Learning Techniques Utilizing Transfer Learning and Layer Freezing Strategies
No full textIn the circulatory system, the presence of embolic particles, which are larger than the red blood cells, is one of the major causes of stroke. Hence, early and reliable detection of these particles is crucial in preventing potential adverse outcomes. Therefore, in this proposed study, 400 Doppler ultrasound signals that belong to three different classes (Speckle, Artifact, and Embolic) are examined for the detection of embolic signals (ESs). Each signal is transformed into a spectrogram image by using the short time Fourier transform, and the proposed learning models are fed by these images called spectrograms. The proposed architecture is developed as a fusion of 10 pretrained Convolutional neural network models, in which the transfer learning and freezing layer approaches are employed. In the fusion of models, the soft and hard voting methods are utilized as the ensemble learning approach. The obtained results show promising performance, achieving a classification accuracy of up to 96.73% and an F1 score of 96.5%. The findings of the study reveal that the proposed ensemble architecture has a high contribution in enhancing the detection of ESs, offering significant implications for stroke prevention strategies
