Bilgisayar Mühendisliği Bölümü Koleksiyonu

Permanent URI for this collectionhttps://hdl.handle.net/20.500.12573/203

Browse

Browsing Bilgisayar Mühendisliği Bölümü Koleksiyonu by WoS Q "N/A"

Filter results by typing the first few letters
Now showing 1 - 3 of 3
  • Thumbnail Image
    Conference Object
    Evaluation of Hybrid Classification Approaches: Case Studies on Credit Datasets
    (Springer Verlag service@springer.de, 2018) Cetiner, Erkan; Güngör, Vehbi Çağrı; Kocak, Taskin; 0000-0003-0803-8372; AGÜ, Mühendislik Fakültesi, Bilgisayar Mühendisliği Bölümü; Gungor, Vehbi Cagri; 01. Abdullah Gül University
    Hybrid classification approaches on credit domain are widely used to obtain valuable information about customer behaviours. Single classification algorithms such as neural networks, support vector machines and regression analysis have been used since years on related area. In this paper, we propose hybrid classification approaches, which try to combine several classifiers and ensemble learners to boost accuracy on classification results. We worked with two credit datasets, German dataset which is a public dataset and a Turkish Corporate Bank dataset. The goal of using such diverse datasets is to search for generalization ability of proposed model. Results show that feature selection plays a vital role on classification accuracy, hybrid approaches which shaped with ensemble learners outperform single classification techniques and hybrid approaches which consists SVM has better accuracy performance than other hybrid approaches. © 2018 Elsevier B.V., All rights reserved.
  • Thumbnail Image
    Conference Object
    Citation - Scopus: 1
    Man-Hour Prediction for Complex Industrial Products
    (Institute of Electrical and Electronics Engineers Inc., 2023) Unal, Ahmet Emin; Boyar, Halit; Kuleli Pak, Burcu Kuleli; Cem Yildiz, Mehmet; Erten, Ali Erman; Güngör, Vehbi Çağrı; 0000-0003-0803-8372; AGÜ, Mühendislik Fakültesi, Bilgisayar Mühendisliği Bölümü; Gungor, Vehbi Cagri; 01. Abdullah Gül University
    Accurately predicting the cost is crucial for the success of complex industrial projects. There can be several sources contributing to the cost. Traditional methods for cost estimation may not provide the required accuracy and speed to ensure the success of the project. Recently, machine learning techniques have shown promising results in improving cost estimation in various industrial products. This study investigates the performance of gradient-boosting machine learning models and feature engineering techniques on a private dataset of metal sheet project man-hour costs. A comparison of distinct models is conducted, key aspects influencing cost are identified, and the implications of incorporating domain-specific knowledge, including its advantages and disadvantages, are assessed based on performance outcomes. Experimental results demonstrate that LightGBM and XGBoost outperform other models, and feature selection and synthetic data generation techniques improve the performance. Overall, this study highlights the potential of machine learning in metal sheet sampling projects and emphasizes the importance of feature engineering and domain expertise for better model performance. © 2024 Elsevier B.V., All rights reserved.
  • Thumbnail Image
    Conference Object
    Citation - Scopus: 2
    Street Vendor Detection: Helping Municipalities Make Decisions With Actionable Insights
    (IEEE, 2021) Agba, Hatice Nur; Tahir, Abdullah; AGÜ, Mühendislik Fakültesi, Bilgisayar Mühendisliği Bölümü; Agba, H.N.; Tahir, A.; 01. Abdullah Gül University
    Street vendors are quite common in countries across the world. By the prevalence of mobile surveillance systems, increasing demand for automatic detection of street vendors for further decisions and planning by the city administrators emerged. In this paper, an object detector is developed using a MobileNet SSD object detection algorithm to detect vendors on the street. For this study images were used, however, in the future this technique could be used for real time video footage from street cameras. Since this is the first study tackling this issue, a data set was created from scratch. The accuracy achieved by the algorithm is promising considering the size of the data set and the minimal computational power available. The goal of this research is to pave the way for more work to be done in this area and help municipalities improve their decision making process regarding street vendor activities in countries like Mexico, Pakistan, China, Turkey, etc.