2018 Cilt 23 Sayı 3

Permanent URI for this collectionhttps://hdl.handle.net/11452/12497

Browse

Browsing by BUU Author "Özcan, Gıyasettin"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    A comparative study for hyperspectral data classification with deep learning and dimensionality reduction techniques
    (Bursa Uludağ Üniversitesi, 2018-10-16) Ortaç, Gizem; Özcan, Gıyasettin; Mühendislik Fakültesi; Bilgisayar Mühendisliği Bölümü
    In recent years, hyperspectral imaging has been a popular subject in the remote sensing community by providing a rich amount of information for each pixel about fields. In general, dimensionality reduction techniques are utilized before classification in statistical pattern-classification to handle high-dimensional and highly correlated feature spaces. However, traditional classifiers and dimensionality reduction methods are difficult tasks in the spectral domain and cannot extract discriminative features. Recently, deep convolutional neural networks are proposed to classify hyperspectral images directly in the spectral domain. In this paper, we present comparative study among traditional data reduction techniques and convolutional neural network. The obtained results on hyperspectral image data sets show that our proposed CNN architecture improves the accuracy rates for classification performance, when compared to traditional methods by increasing the classification accuracy rate by 3% and 6%.
  • Thumbnail Image
    Item
    A new automata based approximate string matching approach and web interface for bioinformatics algorithms
    (Bursa Uludağ Üniversitesi, 2018-10-16) Koca, Burak; Özcan, Gıyasettin; Mühendislik Fakültesi; Bilgisayar Mühendisliği Bölümü
    In this study, we present a new web interface for major bioinformatics algorithms and introduce a novel approximate string matching algorithm. Our web interface executes major algorithms on the field for the use of computational biologists, students or any other interested researchers. In the web interface, algorithms come under three sections: Sequence alignment, pattern matching and motif finding. In each section, we introduce algorithms in order to find best fitting one for specific dataset and problem. The interface introduces execution time, memory usage and context specific results of algorithms such as alignment score. The interface utilizes emerging open source languages and tools. In order to develop light and user-friendly interface, all parts of the interface coded with Python language. On the other hand, Django is used for web interface. Second contribution of the study is novel A-BOM algorithm, which is designed for approximate pattern matching problem. The algorithm is approximate matching variation of Backward Oracle Matching. We compare our algorithm with popular approximate string matching algorithms. Results denote that A-BOM introduces %30 to %80 short runtime improvement when compared to current approximate pattern matching algorithms on long patterns.