Matematik ve Bilgisayar Bölümü / Department of Mathematics and Computer Science
Permanent URI for this collectionhttps://hdl.handle.net/11413/6787
Browse
Browsing Matematik ve Bilgisayar Bölümü / Department of Mathematics and Computer Science by Type "Book chapter"
Now showing 1 - 4 of 4
- Results Per Page
- Sort Options
Publication Blockchain based secure communication for Iot devices in smart sities(2019-04) Yetiş, R.; ŞAHİNGÖZ, ÖZGÜR KORAYIn smart city technologies we have witnessed advanced technological improvements in small computing devices, which can be connected to the Internet and named as Internet of Thing (IoT) devices, and cooperatively working complex systems. With this increased use of new technologies, the security problem is becoming more and more important because complex systems lead to unpredictable security vulnerabilities, which result in financial and private information losses. As a recently emerged technology, Blockchain was emerged as an alternative solution to security breaches of a different application environment. In contrast to the central structure used by most systems, it is preferred especially in the area of security by its distributed structure and the cryptographic hash algorithm it uses. Today, structures such as Smart Home, Smart City, Smart Environment and Smart Agriculture, which are created by using IoT are seen as active research areas with more security shortages. The reason for the security weakness in these areas arises from the hardware restriction on the IoT devices used. In the proposed system, an authorization system for IoT devices has been tried to be set up by using the distributed node structure of Blockchain system and blocks kept in these nodes. UDP (User Datagram Protocol), which uses a simple communication model without establishing a connection to the minimum protocol mechanism for communication of nodes in the system, was preferred. The communication between the nodes has been encrypted using encryption methods, thus creating a secure environment.Publication Deep learning based classification of malaria from slide images(2019) Kalkan, Soner Can; ŞAHİNGÖZ, ÖZGÜR KORAYAs one of the most life-threatening disease in the tropical and warmer-climate countries, Malaria affects not only animals but also humans who can be infected by only a single bite from a mosquito. Although this disease is wiped out in high-income countries, as a result of traveling people, it can even emerge in all part of the world. World Health Organization announced that more than 400,000 people are expected to die due to this illness. However, it is a curable and preventable disease, if early detection is possible. Traditionally, Pathologists diagnosed this disease manually by using microscope which is a time-consuming process in our computerized world, and this model depends on the experience of the Pathologists, which is a critical problem in rural areas. Therefore, in recent years detection of Malaria using computerized image analysis which is trained using some dynamic learning mechanism has gained increasing importance. In this paper, we proposed an image processing-based Malaria detection system which is trained by deep learning. We used relatively big data for increasing the accuracy of the system, and the reached accuracy showed that the proposed system has an outstanding classification rate that can be used in real-world detection.Publication Diffused label popagation based transductive classification algorithm for Wwrd sense disambiguation(2019-07) Kocaman, G.; Gerek, A.; Altınel, B.; Ganiz, M.C.; ŞİPAL, BİLGEA major natural language processing problem, word sense disambiguation is the task of identifying the correct sense of a polysemous word based on its context. In terms of machine learning, this can be considered as a supervised classification problem. A better alternative can be the use of semi-supervised classifiers since labeled data is usually scarce yet we can access large quantities of unlabeled textual data. We propose an improvement to Label Propagation which is a well-known transductive classification algorithm for word sense disambiguation. Our approach make use of a semantic diffusion kernel. We name this new algorithm as diffused label propagation algorithm (DILP). We evaluate our proposed algorithm with experiments utilizing various sizes of training sets of disambiguated corpora. With these experiments we try to answer the following questions: 1. Does our algorithm with semantic kernel formulation yield higher classification performance than the popular kernels? 2. Under which conditions does a kernel design perform better than others? 3. What kind of regularization methods result with better performance? Our experiments demonstrate that our approach can outperform baseline in terms of accuracy in several conditions.Publication Smart home hecurity with the use of WSNs on future intelligent cities(2019-04) Dine, G.; ŞAHİNGÖZ, ÖZGÜR KORAYIn recent years, smart cities use some advanced technologies such as Cyber-Physical Systems and Internet of Things for improving the quality of human life by enabling extra security and communication abilities. In a recent survey of ICMA which is the world's leading association of professional city and county managers, ICMA emphasizes the importance of the safety and security benefits of smart cities by categorizing it as one of the five important motivating factors. Therefore, in this paper, we proposed a new security model for smart buildings and smart homes by using a hot topic research area of Cyber-Physical Systems: WSNs. With the use of this system, it is aimed to construct a smart security system by using a minimum number of sensor nodes with the increased coverage rate for smart buildings and homes.