Master Degree / Yüksek Lisans Tezleri

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

Browse

Filters

Settings

Department: search.filters.department.Thesis (Master)--İzmir Institute of Technology, Computer EngineeringPublisher: search.filters.publisher.Izmir Institute of Technology

Search Results

Now showing 1 - 10 of 148
  • Master Thesis
    Development of Co-Evolution Tracker Tool for Software With Acceptance Criteria
    (Izmir Institute of Technology, 2022) Yalçın, Ali Görkem; Tuğlular, Tuğkan; Tuğlular, Tuğkan
    Testing is a vital part of achieving good-quality software. Deploying untested code can cause system crashes and unexpected behavior. In order to reduce these problems, testing must be prioritized. However, once test suites are created, they should not remain static throughout the software updates. Since whenever software gets updated, new functionalities are added or existing functionalities are changed, so whenever the application is updated, test suites must be updated along with the software. If the old test suites are used with the new updates, unexpected testing results can occur. In order to repair test cases in the process of software evolution, analyzing real-world projects’ software and test case evolution is an important prerequisite. Software repositories contain valuable information about the software systems. Having access to older versions and by differentiating adjacent versions’ test and production code changes can provide information about the evolution process of the software. This thesis concentrates on the development of a tool that is used for the analysis of 21 real-world projects in the terms of co-evolution of both software and its test suites. Related projects are retrieved from repositories and filtered according to this study’s needs, then for each project's every update is analyzed, and graphs and analysis related to the co-evolution process are created.
  • 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.
  • Master Thesis
    Compiler-Managed Fault Tolerance Techniques for General Purpose Graphics Processing Units
    (Izmir Institute of Technology, 2022) Kaya, Ercüment; Öz, Işıl; Öz, Işıl
    As the use of graphics processing units evolves for general-purpose computations besides inherently-fault tolerant graphics programs, soft error reliability becomes a first-class citizen in program design. In this thesis, we aim to increase the reliability of general-purpose graphics processing units. We propose compiler-based redundancy schemes for graphics processing units. Our framework replicates the annotated kernel function by a programmer at compile time. Our selective redundancy approach enables us to provide full redundancy with no error and partial redundancy with an acceptable error rate with higher performance. We develop different schemes to satisfy the performance and memory requirements of the general-purpose graphics processing unit applications. We build our framework on top of the LLVM compiler framework to increase the reliability of applications that exploit the CUDA programming model and evaluate our schemes for the applications from the PolyBench benchmark suite. We reveal that our partial redundancy approach improves the reliability with a small performance overhead and our full redundancy schemes provide complete fault coverage with varying performance differences based on the application's characteristics.
  • Master Thesis
    A Blockchain Application for Payment and Traffic Management in Smart Vehicles
    (Izmir Institute of Technology, 2022) Yiğitbaşı, Boğaçhan; Ayav, Tolga; Ayav, Tolga
    The proposed solution offers an alternative way to our current retail shopping of fuel fees. It can be applied to any retail shopping process but this phase of the project is considered an initiation of upcoming. The next phases of the project, include full integration with smart cars in order to handle all procedures automatically. In the traditional way, when you buy some gas from a station with your credit card, the station owner pays some fee to his bank and it has to wait for some time to be able to get that money. You as an individual have to expose your identity so they can track your shopping habits and follow your expenses. Sometimes they may offer some loyalty discounts or gifts but with really ridiculous rates. Our system offers a digital payment system based on the Ethereum blockchain. It has its own token called TRANT (Transport Token) and by this token, any client with a digital wallet (Metamask) is able to pay their gas fees without exposing their real identity -only their wallet address-, and get some rewards in terms of TRANT for their loyalties and using our DEX (Decentralized Exchange) exchange them into the ether which can be converted into real fiat money easily. On the other hand, the proposed solution also has some advantages for the other party in this equation such as gas station owners, they immediately get their money at that very first moment without any remittance.
  • Master Thesis
    Design and Implementation of a Domain Specific Language for Event Sequence Graphs
    (Izmir Institute of Technology, 2022) Kalecik, Mert; Tuğlular, Tuğkan
    Nowadays, large-scale software applications are being developed because of the increasing q-commerce or e-commerce conversion rate. Companies extend their service operation areas with the trend of having a super app. As the result of extended functionality brings some risks together. Therefore, software quality is one of the crucial metrics for achieving reliable and faultless software products. One way of achieving software quality is systematic testing, which is often materialized by model-based testing. An example of model-based testing approaches is Event Sequence Graphs (ESGs). Domain specific language is usually a declarative language that provides substantial gain on a restricted business domain. This thesis mainly focuses on the development of a domain specific language (DSL) for ESG building and visualization process with a modularization support for sub-ESGs and decision tables. The ESGs are augmented by decision tables visualized with a vertex and that vertex is visualized with two tables such as property table and property definition table. The use of the proposed DSL is compared with the existing ESG tool called Test Suite Designer (TSD) in areas such as measuring the cost of quality, understanding the value of quality, motivation to achieve quality, and understand how to overcome it. The comparison results obtained through a questionnaire applied to a focus group show that some improvements for both ESG DSL and TSD are necessary.
  • Master Thesis
    Wirelless Mesh Network Throughput Analysis Using Petri Nets
    (Izmir Institute of Technology, 2022) Oğuzer, Lütfü Melih Buğra; Tuğlular, Tuğkan; Belli, Fevzi
    Evolving technology has made the understanding of quality perception in software processes more difficult. Unlike other sectors, rapid adaptation and software development processes have become a critical issue. This issue can especially be observed in the service, telecommunication, and high technology sectors. User demands and competition are quite high and with this competition, the need to subject the customized or developed software to rapid testing processes has formed. Undoubtedly, this process implies a great responsibility for the "quality assurance" teams. This responsibility has reached a level that can only be handled by the quality assurance departments that automate the testing cycles. However, it is also important that these cycles are very efficient. Our research is concerned with modeling test processes with Petri nets and creating test scenarios based on this modeling to make automation processes in the telecommunications industry more efficient. In this research, the performance analysis of wireless mesh networks is executed through place/transition petri-net modeling. Through this modeling, reusable test scenarios which were compared and analyzed with traditional automation processes were created for performance tests. The research also addresses another topic which is the shortening of the modeling processes created with Petri nets and how to make them more efficient. In this context, a tool has been developed in order to shorten the modeling process and analyze the reusable test scenarios. Finally, ten test engineers were interviewed about reusable test processes. In these interviews, feedback was provided on reusable test scenarios in test automation processes.
  • Master Thesis
    Evaluation of Scheduling Architectures for Osek/Vdx Compliant Hard Real-Time Operating Systems
    (Izmir Institute of Technology, 2020) Saydam, Berkay; Ayav, Tolga
    Technological advancements are reflected to the vehicles as well, but it brings the challenge of adding new functionalities to vehicles without compromising safety. Tasks are used to provide functionalities which are used in car. These tasks have different characterictics. Safety and performance are two main criteria to determine characterictic of tasks. Characteristics of tasks can be classified according to their safety levels which are known as Automotive Safety Integrity Levels. Designing of hardware and software and also testing them is a long progress in automotive industry. Any changes on the design of hardware is quite costly when an ECU began to be used in field. According to my hypothesis, scheduling algorithms which are used by Central Processing Unit to determine sequences of task executions, should be well known. Besides, designing of hardware and software should be done according to these characteristics and algorithms. If not, tasks will cause serious problem like missing deadline for safety-critical component. In this thesis, the scheduling architectures are evaluated and they are determined which scheduling architectures should be used for which purpose. Besides, the advantages and disadvantages are explained.
  • Master Thesis
    Synthetic Generation of Fingerprints
    (Izmir Institute of Technology, 2020) İrtem, Emre; Erdoğmuş, Nesli
    Fingerprints are unique to each person and they have been widely used and accepted for identification purposes by the society. Fingerprints can be captured by using ink and paper to get a print and then digitizing it or more recently by using specialized sensors. But in both cases, trained specialist supervision is mostly needed. Moreover, since fingerprints are personal information, they are protected by the laws on personal data protection. Therefore, collection/sharing of real fingerprints is difficult and illegal without the consent of their owner. On the otherhand, deep learning systems that are proven to be very successfull in many machine learning task, usually depend on very large training sets to achive high accuracies. In this study, to overcome the data hunger problem for training deep neural networks, synthetic fingerprints are generated by using model-based methods. For this purpose, firstly master fingerprint images are generated and next many impressions are derived from them by applying real-world degradations. The realism and the usability of synthetic fingerprints are tried and validated using a fingerprint classification system. For which, a deep neural networks are trained with and without the synthetically generated data. As a result of the experiments, it is shown that the generated fingerprint images are realistic enough to positively effect the classification results and that the usage of the synthetically generated fingerprints in training deep systems are promising.
  • Master Thesis
    A Language Modeling Approach To Detect Bias
    (Izmir Institute of Technology, 2020) Atik, Ceren; Tekir, Selma
    Technology is developing day by day and is involved in every area of our lives. Technological innovations such as artificial intelligence can strengthen social biases that already exist in society, regardless of the developers' intentions. Therefore, researchers should be aware of this ethical issue. In this thesis, the effect of gender bias, which is one of the social biases, on occupation classification is investigated. For this, a new dataset was created by collecting obituaries from the New York Times website and they were handled in two different versions, with and without gender indicators. Since occupation and gender are independent variables, gender indicators should not have an impact on the occupation prediction of models. In this context, in order to investigate gender bias on occupation estimation, a model in which occupation and gender are learned together is evaluated as well as models that make only occupation classification are evaluated. The results obtained from models state that gender bias has a role in classification occupation.
  • Master Thesis
    Comparison of Classification Algorithms in Pitch Type Prediction Problem
    (Izmir Institute of Technology, 2020) Türkmen, Fatih; Ergenç Bostanoğlu, Belgin
    The dramatic increase in the use of IoT devices has been leading to a huge amount of valuable data to be discovered. The knowledge extraction from such a huge amount of data requires an organized scientific set of processes. This requirement has pointed out the importance of the data mining applications. As a major data mining application, classification is a supervised learning technique that requires a feature set and target class through the training process. For the training process, the key point is determining the appropriate feature set for the classification algorithm. The improvements in cutting-edge technologies such as high resolution camera systems have made extracting the insights about next pitch available. Consequently, pitch type prediction has been standing out as an important research topic. In order to predict next pitch type, existing researches mostly focus on pitcher profile, batter profile and previous pitch data in feature set. There is no study analyzing the effect of the zone information in the prediction of the next pitch type. Therefore, this study has analyzed the contribution of zone information in pitch type prediction. Our approach is that, we aimed to reveal the contribution of zones with the high strike low bat rates for pitch type decision in pitcher and batter player match up. This aim directed us to analyze the pitch type prediction problem for both zone-based and non-zone-based approaches so that we can exhibit how much zone information contributes to the problem through different classification algorithms.