Master Degree / Yüksek Lisans Tezleri

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

Browse

Filters

2022 - 20233

Settings

Department: search.filters.department.Thesis (Master)--İzmir Institute of Technology, Computer EngineeringSubject: search.filters.subject.Data sets

Search Results

Now showing 1 - 3 of 3
  • Master Thesis
    Recognition of Counterfactual Statements in Turkish
    (01. Izmir Institute of Technology, 2023) Acar, Ali; Tekir, Selma
    Counterfactual statements describe an event that did not happen or cannot happen, and optionally the consequence of this event if it would happen. Counterfactual statements are the building blocks of human thought processes as people constantly reflect upon past happenings and consider their future implications. Counterfactual reasoning is essential for machine intelligence and explainable artificial intelligence studies. Detecting counterfactuals automatically with machine learning algorithms is very crucial for these areas. This thesis presents the development of the first-ever Turkish counterfactual detection dataset. It presents a comprehensive classification baseline and expands the scope of counterfactual detection to include the Turkish language.
  • Master Thesis
    A Mutation-Based Approach To Alleviate the Class Imbalance Problem in Software Defect Prediction
    (01. Izmir Institute of Technology, 2023) Güner, Dinçer; Demirörs, Onur; Demirörs, Onur; Giray, Görkem
    Highly imbalanced training datasets considerably degrade the performance of software defect predictors. Software Defect Prediction (SDP) datasets have a general problem, which is class imbalance. Therefore, a variety of methods have been developed to alleviate Class Imbalance Problem (CIP). However, these classical methods, like data-sampling, balance datasets without connecting any relation with SDP. Over-sampling techniques generate synthetic minor class instances, which generalize a small number of minor class instances and result in less diverse instances, whereas under-sampling techniques eliminate major class instances, resulting in significant information loss. In this study, we present an approach that uses software mutations to balance software repositories. Mutation-based Approach (MBA) injects mutants into defect-free instances, causing them to transform into defective instances. In this way, MBA balances datasets with diverse data produced by mutation operators, and there is no loss on instances as in under-sampling. For recall scores, almost all rebalancing methods outperformed Baseline in Inter-release Defect Prediction (IRDP) scenario but only MBA significantly outperformed Baseline in Cross-project Defect Prediction (CPDP) scenario. The performance increase in recall resulted in the production of more false alarms. We can not generalize that MBA outperforms Baseline and the five over-sampling strategies in terms of AUC scores. In terms of recall values, the MBA performed better in CPDP than IRDP. For both IRDP and CPDP scenarios, there were significant and positive correlations between SMC (the change percentage of software measures) and recall, and SMC and false alarm but there was no significant correlation between SMC and AUC.
  • Master Thesis
    Hierarchical Image Classification With Self-Supervised Vision Transformer Features
    (Izmir Institute of Technology, 2022) Karagüler, Caner; Özuysal, Mustafa
    There are lots of works about image classification and most of them are based on convolutional neural networks (CNN). In image classification, some classes are more difficult to distinguish than others because of non-even visual separability. These difficult classes require domain-specific classifiers but traditional convolutional neural networks are trained as flat N-way classifiers. These flat classifiers can not leverage the hierarchical information of the classes well. To solve this issue, researchers proposed new techniques that embeds class-hierarchy into the convolutional neural networks and most of these techniques exceed existing convolutional neural networks' success rates on large-scale datasets like ImageNet. In this work, we questioned if a hierarchical image classification with self- supervised vision transformer features can exceed hierarchical convolutional neural networks. During this work, we used a hierarchical ETHEC dataset and extract attention features with the help of vision transformers. Using these attention features, we implemented 3 different hierarchical classification approaches and compared the results with CNN alternative of our approaches.