Bilgisayar ve Bilişim Fakültesi Koleksiyonu

Permanent URI for this collectionhttps://hdl.handle.net/20.500.13091/10834

Browse

Now showing 1 - 8 of 8

Classification of Sleep Stages Using Psg Recording Signals
(2020) Koca, Yasin; Özşen, Seral; Göğüş, Fatma Zehra; Tezel, Gülay; Küççüktürk, Serkan; Vatansev, Hülya
Automatic sleep staging is aimed within the scope of this paper. Sleep staging is a study by a sleep specialist. Since this process takes quite a long time and sleep is a method based on the knowledge and experience, it is inevitable for each person to show different results. For this, an automatic sleep staging method has been introduced. In the study, EEG (Electroencephalogram), EOG (Electrooculogram), EMG (Electromyogram) data recorded by PSG (Polysomnography) device for seven patients in Necmettin Erbakan University sleep laboratory were used. 81 different features were taken from the data in time and frequency environment. Also, PCA (Principal component analysis) and SFS (Sequential forward selection) feature selection methods were used. The classification success of the sleep phases in different machine learning methods was measured by using the received features. Linear D. (Linear Discriminant Analysis), Cubic SVM (Support vector machine), Weighted kNN (k nearest neighbor), Bagged Trees, ANN (Artificial neural network) were used as classifiers. System success was achieved with a 5 fold cross-validation method. Accuracy rates obtained were respectively 55.6%, 65.8%, 67%, 72.1%, and 69.1%.
Etkin Eeg Özellikleri Çıkarılarak Arousal Tespiti
(2020) Erol, Gizemnur; Göğüş, Fatma Zehra; Tezel, Gülay
Son zamanlarda toplumun en önemli problemlerinden biri olan uyku bozuklukları, bireylerin sağlığını ve yaşam kalitesini ciddi şekilde etkilemektedir. Uykusuzluk (Insomnia), narkolepsi, uyku apnesi ve huzursuz bacak sendromu gibi birçok uyku bozukluklarının neden olduğu rahatsızlıklar vardır. Uyku bozukluklarına sebep olan ana faktör ise bireyin uyku anındaki uyanma ile sonuçlanamayan, uyku kalitesini düşüren uyku kesintileridir. Arousal diğer bir adı ile uyanayazma geçici olan bu kesintilerdir ve bir beyin dalga (Elektroansefalogram -EEG) aktivitesinin paternindeki ani değişikliği temsil etmektedir. Arousal tespiti genellikle EEG verileri kullanılarak Amerikan Uyku Tıbbı Akademisi (American Academy of Sleep Medicine-AASM) tarafından belirlenen kriterlere göre yapılmaktadır. Bu çalışmada amaç, AASM tarafından belirlenen kriterler doğrultusunda EEG sinyalleri vasıtasıyla hasta bireylerdeki arousalların tespitidir. Bu amaç doğrultusunda, öncelikle, çalışmaya dahil edilen 5 hasta bireyin tek kanallı (C3/A2) EEG sinyallerine sırasıyla filtreleme, normalizasyon ve segmantasyon önişlemleri uygulanmıştır. Daha sonra Spektral Güç Yoğunluğu (Power Spectral Density-PSD) ve Ayrık Dalgacık Dönüşümü (Discrete Wavelet Transform-DWT) yöntemleri ile gerçekleştirilen özellik çıkarma süreci sayesinde, EEG sinyal segmentlerine ait 2 özellik seti ve bu özellik setlerinin birleştirilmesiyle 3. özellik seti oluşturulmuştur. Ardından oluşturulan 3 özellik seti üzerine Sarmal Alt Küme Değerlendirme (Wrapper Subset Evaluation-WSE) özellik seçme yöntemi uygulanarak etkin özellikler belirlenmiştir. Nihai olarak belirlenen özelliklerin Yapay Sinir Ağları (YSA) ve Rasgele Orman (RO) algoritmaları tarafından sınıflandırılmaları ile arousal içeren EEG segmentleri tespit edilmiştir. Gerçekleştirilen bu çalışmaların beraberinde EEG sinyal kayıtlarından başka hiçbir PSG sinyal kaydına ihtiyaç duymadan, yalnızca tek kanallı EEG sinyalleri ile oldukça başarılı sonuçlar elde edildiği tespit edilmiştir. Çalışma sonucunda ise Özellik Seti 3’ün etkin özellikleri ve YSA ile en yüksek doğruluk oranı %99.05 olarak elde edilmiştir.
Geri Dönüştürülebilir Atıkların Materyallerine Göre Sınıflandırılması için Raspberry Pi Tabanlı Donanım Geliştirilmesi
(2020) Sağlam, Ali; Taş, Melike; Baykan, Nurdan Akhan
Hem doğanın korunması hem de sürekli artan insan ihtiyaçları için gerekli olan ve doğada kısıtlı miktarda bulunan materyallerin takviye edilmesi için ortaya çıkan “geri dönüşüm” kavramı son yıllarda en önemli konulardan birisi olmuştur. Belirli bir geri dönüşüm işlemi sonucunda, “ham maddesi yeniden kullanılabilir hale getirilebilen atıklar” olarak bilinen geri dönüştürülebilir atıkların toplanması konusu dünya genelinde üst ve yerel yönetimlerin de ilgilendiği bir problem olmuştur. Bunun için belirli merkezlere geri dönüştürülebilir atıklar için özel kutular yerleştirilmekte ve insanlar geri dönüşüm konusunda teşvik edilmeye çalışılmaktadır. Bu çalışmada, geri dönüşüm projelerinde kullanılmak üzere kâğıt, cam ve plastik atıklarının geri dönüşüm kutuları içerisinde gerçek zamanlı olarak tespit edilebilmesi için gerekli elektronik malzemeler ve yazılımlar kullanılarak bir materyal tanıma sistemi geliştirilmektedir. Sistem geri dönüşüm kutusuna atılan geri dönüştürülebilir katı atıkların materyallerini tanıyan ve materyale göre kullanıcı hesabına ücret yükleyen bir simülasyon işlevi görmektedir. Geliştirilen donanım kamera, LCD ekran, LED, IR LED, devre tahtası ve jumper kablo gibi Raspberry Pi üzerine bağlanabilen elektronik cihazları da içermektedir. Materyallerin tanınması için gerekli yazılımının geliştirilmesi aşamasında; kâğıt, cam ve plastik materyallerini içeren 845 adet resim çalışma kapsamında hazırlanmış ve bunların 662 tanesi Tensorflow nesne tanıma kütüphanesi üzerinde eğitim için kullanılmıştır. Materyallerin geliştirilen donanım tarafından gerçek zamanlı olarak algılanması ve elde edilen nesne tanıma modelinin donanım üzerinde kullanılabilmesi için Raspberry Pi içerisine OpenCV bilgisayarlı görme kütüphanesi yüklenmiştir. En son olarak, geliştirilen donanım ilgili materyallere özel ayrılmış kutular üzerine sabitlenerek sistem gerçek zamanlı olarak çalışır hale getirilmiştir. Sistemin düzgün çalıştığını doğrulamak için kutu içerisine bazı atıklar atılmış ve LCD ekran üzerinde sonuçlar görüntülenmiştir.
Modified Region Growing Method for Image Segmentation Using Ant Lion Optimization Algorithm
(2020) Jama, Bashir Sheikh Abdullahi; Akhan Baykan, Nurdan
Image segmentation is a significant step in image processing that applies to various fields. These fields include machine vision, object detection, astronomy, biometric recognition systems (face, fingerprint, plate, and eye), medical imaging, video surveillance, and many other image-based technologies. Efficient image segmentation is one of the most important tasks and critical roles in automatic image processing. Especially in engineering studies, finding the most suitable solutions for problems is one of the important research topics. Bio-inspired algorithms such as Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Artificial Bee Colony (ABC), and Bat Algorithm (BAT), etc. are used to find the optimal solutions in search spaces and Ant Lion Optimization (ALO) is one of these algorithms. In recent years, bio-inspired algorithms are used to optimize the segmentation parameters of the images. This research proposes a modified region growing (RG) image segmentation approach using bio-inspired ALO. Region growing (RG) has three main problems as the selection of the right seeds, the number of seeds, and the region growing strategy. Therefore, ALO was used to solve seed selection problems in RG. In this study, firstly, the median filter was applied to the inputs to improve the quality of the images. Subsequently, the region growing segmentation was carried out using optimal seed points obtained from the ALO. For obtaining the optimal seeds, ALO was used to solve the limitations of RG during the segmentation process. The success of the proposed approach was tested using some images taken from the BSDS300 (Berkeley) dataset. The experimental results show that the proposed method segments almost all the images.
Network Intrusion Detection Using Optimized Machine Learning Algorithms
(2021) Khorram, Tahira; Baykan, Nurdan Akhan
Network intrusion detection mechanism is a primary requirement in the current fast-growing network systems. Data mining and machine learning approaches are widely used for network anomaly detection during past few years. Machine learning based intrusive activity detector is becoming more popular. The most commonly used machine learning algorithms for Intrusion Detection System (IDS) are K-Nearest Neighbor (KNN), Support Vector Machine (SVM) and Random Forest (RF). However, the performance of these methods is reliant upon the selection of the proper parameter values. This research focuses its aim to build an IDS model based on the most effective algorithms. The machine learning algorithms are used in this research are KNN, SVM and RF. To improve these algorithms classification accuracy, some parameters of the algorithms are optimized using Particle Swarm Optimization (PSO) and Artificial Bee Colony (ABC) optimization techniques, while other parameters are used with default values. The result of this experiment shows that optimized KNN, SVM and RF perform better than these algorithms with their default parameter values. Furthermore, the results the experiment shows that KNN is the most suitable algorithm for network anomaly detection regarding detection of known network attacks and unknown network attacks. NSL-KDD standard dataset is used for the experiments of this research. It has been proven that our proposed model performs better than what is provided in the state-of-arts models.
A Novel Hybrid Iot Based Ids Using Binary Grey Wolf Optimizer (bgwo) and Naive Bayes (nb)
(2020) Nur, İsmail Mohamed; Ülker, Erkan
One of the main objectives of intelligent environments is to enhance the quality of human life standard in terms of efficiency and comfort. The Internet of Things (IoT) model has newly evolved into the technology for establishing smart environments. IoT refers to physical things or devices which are able to exchange information with other devices. It is used in various fields such as smart home, smart city, industrial control, automobile industry, agriculture, intelligent transportation, home automation and appliances, healthcare, and many other fields. Moreover, it assures innovative business paradigms and advanced user experience. Privacy and security are counted as the key problems in any real-world intelligent environment for the IoT paradigm. Therefore, to implement the security of the IoT systems is becoming the first priority and big area of interest in the successful distribution of IoT networks. The open holes of security in IoT related systems create security risks that impact the smart applications. Mirai botnet is an example of one of the novel attacks that launched recently. The network of IoT is protected with authentication and encryption, but it can’t be mitigated against malicious and harmful attacks. Thus, IoT based Intrusion Detection System (IDS) is required to detect the attacks. In this paper, a novel hybrid IoT based IDS using Binary Grey wolf optimizer (BGWO) and Naive Bayes (NB) is presented to defend and secure intrusions on the IoT network. BGWO is used as feature selection and NB as a classification method. The results are compared with other optimization algorithms. The BoT-IoT data set is used as an experimental data set.
Nsga-Ii Algorithm for the Reallocation Problem in Land Consolidation
(2020) Ortaçay, Zeynep; Uğuz, Harun; Haklı, Hüseyin
To solve problems encountered in real life, we sometimes need optimization algorithms. Some of these problems have single objective, while others have multiple objectives. If there is a single objective, the problem is defined as a single-objective optimization problem and if there are more than one objective it is called multi-objective optimization problem. Today, lands are fragmented and scattered. This makes agriculture difficult and costly. To prevent these problems, Land Consolidation (LC) studies are being carried out. The reallocation stage, which is part of LC, can be defined as a multi objective optimization problem. In this study, one of the multi objective optimization techniques, NSGA-II algorithm, is applied to the reallocation problem. The results are comparable with the studies in the literature.
Prediction of Diabetes Mellitus by Using Gradient Boosting Classification
(2020) Nusrat, Fatema; Uzbaş, Betül; Baykan, Ömer Kaan
Diabetes has become a pervasive and endemic health problem worldwide. It is a chronic disease and also life-threatening. It can cause health problems in many organs such as the heart, kidneys, eyes, nerves, and blood vessels. To reduce the fatality rate from diabetes, early prevention techniques are needed. Nowadays, machine learning techniques are used to predict or detect different life-threatening diseases like cancer, diabetes, heart diseases, thyroid, etc. In this study, a prediction model of diabetes mellitus was presented using the Pima Indian dataset. Three different machine learning techniques that Decision Tree (DT), Random Forest (RF) and, Gradient Boosting (GB) algorithm were used to predict diabetes mellitus and the performance analysis was performed. Confusion matrix, accuracy, F1 score, precision, recall, Cohen’s kappa were evaluated and also a ROC curve was plotted. Out of the three techniques, the best results have been achieved with GB.

Browse

Browsing Bilgisayar ve Bilişim Fakültesi Koleksiyonu by Journal "Avrupa Bilim ve Teknoloji Dergisi"