Master Degree / Yüksek Lisans Tezleri

Permanent URI for this collectionhttps://hdl.handle.net/11147/3008

Browse

Filters

2018 - 201922020 - 20222

Settings

Department: search.filters.department.Thesis (Master)--İzmir Institute of Technology, Computer EngineeringSubject: search.filters.subject.Artificial neural networks

Search Results

Now showing 1 - 4 of 4
  • Master Thesis
    Development of a Machine Learning Platform for Analysis of Mitochondrial Features in Live-Cell Images
    (01. Izmir Institute of Technology, 2022) Tarkan, Yalçın; Tuğlular, Tuğkan; Tuğlular, Tuğkan
    It is a laborious and error-prone manual process to mark the organelles in 2D and 3D images of living cells and identify the behavioral feedback to stimulations under measured conditions. This manual process can be simplified by being largely automated with machine learning techniques. We created a machine learning-based software platform named MitoML, which extracts sub-cellular structures, specifically mitochondria, and helps to identify the effects of external factors or changes under natural conditions. We investigate appropriate machine learning techniques for these objectives. Image processing and segmentation techniques with neural networks, enable researchers to carry out experiments with much better accuracy and a larger scale by automatically segmenting and counting the mitochondria, calculate the energy potentials based on region brightness. This way, analysis of mitochondria feedback in healthy and cancer cells under various conditions, such as nanomedicine and different treatment therapies, can be performed using MitoML. As a result of our work, we proposed a cascaded neural network architecture that can identify and count mitochondria, quantify its energy levels in fluorescence and other microscopy images, fast and at a standard reliable accuracy. Our test results outperformed the classical image processing algorithms provided in segmentation tools and software for medical image segmentation which was taken as a base line. Achieved accuracy rates 93.4% and %89.6 according to Dice and IoU metrics respectively are also better than any other related work encountered during the research. The proposed method can be improved to cover other sub-cellular structures relieving the researchers from non-standardized and laborious manual work which is prone to human error.
  • Master Thesis
    Application of Graph Neural Networks on Software Modeling
    (01. Izmir Institute of Technology, 2020) Leblebici, Onur Yusuf; Tuğlular, Tuğkan; Belli, Fevzi
    Deficiencies and inconsistencies introduced during the modeling of software systems can cause undesirable consequences that may result in high costs and negatively affect the quality of all developments made using these models. Therefore, creating better models will help the software engineers to build better software systems that meet expectations. One of the software modelling methods used for analysis of graphical user interfaces is Event Sequence Graphs (ESG). The goal of this thesis is to propose a method that predicts missing or forgotten links between events defined in an ESG via Graph Neural Networks (GNN). A five-step process consisting of the following steps is proposed: (i) data collection from ESG model, (ii) dataset transformation, (iii) GNN model training, (iv) validation of trained model and (v) testing the model on unseen data. Three performance metrics, namely cross entropy loss, area under curve and accuracy, were used to measure the performance of the GNN models. Examining the results of the experiments performed on different datasets and different variations of GNN, shows that even with relatively small datasets prepared from ESG models, predicts missing or forgotten links between events defined in an ESG can be achieved.
  • Master Thesis
    Detection and Localization of Motorway Overhead Directional Signs by Convolutional Neural Networks Trained With Synthetic Images
    (Izmir Institute of Technology, 2019) Hekimgil, Hakan; Baştanlar, Yalın
    Image classification, object detection and recognition have gone a long way in the last decade. The competitions, starting with ImageNet, have shown that various improving implementations of Artificial Neural Networks are the best Machine Learning techniques at the time for such tasks. However, machine learning methods require much training data and the such data for image related tasks come at a cost in terms of time and effort, if it can be obtained at all. When training data is scarce or not representative of the whole target set, synthetic data and data augmentation methods are used to increase the training data using what is already available. This thesis work shows that when the target classification images have a structure, even a loose one, it is still possible to use machine learning methods, deep learning in this case, without any real data to begin with and still produce a good detection model. In this work, a Convolutional Neural Network model is trained to detect and localize informative motorway lane direction signs. Starting with no real samples of the target images, a large computer-generated training set is created to train the model. The resulting detector can detect the required sign types with high accuracy, localizing their position by bounding boxes and categorizing them.
  • Master Thesis
    A Systematic Evaluation of Semantic Representations in Natural Language Processing
    (Izmir Institute of Technology, 2018) Sevgili Ergüven, Özge; Tekir, Selma
    In the studies of semantics, the main aim is to address meaning. In a computational manner, this goal is accomplished through the encoding of language constructs. These encodings are in the form of information-theoretic measures and vector representations. We have focused on the representation of words. In word representations, the earlier approaches depend on counting the statistics between word and its accompanied words, whereas the current methods are based on learning approaches. At this point, we have investigated the relation between these two approaches. We have realized that both approaches use context as the normalization factor. We support our idea by evaluating word representations on some Natural Language Processing (NLP) tasks. Furthermore, we have studied the polysemous words which carry more than one meaning. The word representation of the polysemous word provides a representation that covers more than one meaning. To overcome this issue, we provide a method to create a representation for each sense of polysemous word.