HAKAN AYDIN

Makale | 2022 | Applied Computer Science

Increasing dependence on Information and Communication Technologies (ICT) and especially on the Internet in Industrial Control Systems (ICS) has made these systems the primary target of cyber-attacks. As ICS are extensively used in Critical Infrastructures (CI), this makes CI more vulnerable to cyber-attacks and their protection becomes an important issue. On the other hand, cyberattacks can exploit not only software but also physics; that is, they can target the fundamental physical aspects of computation. The newly discovered RowHammer (RH) fault injection attack is a serious vulnerability targeting hardware on reliability and sec . . .urity of DRAM (Dynamic Random Access Memory). Studies on this vulnerability issue raise serious security concerns. The purpose of this study was to overview the RH phenomenon in DRAMs and its possible security risks on ICSs and to discuss a few possible realistic RH attack scenarios for ICSs. The results of the study revealed that RH is a serious security threat to any computerbased system having DRAMs, and this also applies to ICS Daha fazlası Daha az

HAKAN AYDIN

Makale | 2022 | Academic Journal of Information Tecnology

Dijital platform kullanıcıları, bu platformların sunduğu özelleştirilmiş hizmetlerden yararlanmak ve bunları zaman ve mekan bağımsız olarak tüketmek istemektedirler. İnternet üzerinden yayın yapan bu platformlar arasında dünya çapında en yaygın olanlardan biri de Netflix’tir. Bu çalışmanın amacı TF-IDF (term frequency–inverse document frequency) algoritması ve Kosinüs Benzerliği (Cosine Similarity) ile Doğal Dil İşleme (NLP) ile Netflix kullanıcı verileri üzerinde bir içerik öneri sistemi uygulaması geliştirmektir. Bu bağlamda çalışmamızda yapılan analizler ile benzerlik yöntemleri ve uygun eşleşme verilerinin bulunması, böylelikle . . .kullanıcılara kişisel bazda öneri yapılması hedeflenmiştir. Çalışma kapsamında hem Türkçe hem de diğer dillerdeki filmler ve diziler üzerinde farklı deneyler yapılmıştır. Yapılan deneyler neticesinde kosinüs benzerliği kullanılarak en yüksek benzerlik başarısı %91, en düşük benzerlik başarısı ise %43 olarak elde edilmiştir. Deneyler aynı veriler üzerinde kosinüs benzerliği ile birlikte TF-IDF algoritması ile yapıldığında ise başarı oranı %99 ile %80 arasında elde edilmiştir. Çalışma sonuçları, TF-IDF algoritması ve kosinüs benzerliği birlikte uygulanarak yapılan deneylerde, kosinüs benzerliği kullanılarak yapılan deneylere nazaran daha yüksek başarı oranının elde edildiğini ortaya koymaktadır. Çalışmamızın benzerlik yöntemleri ve uygun eşleşme verileri kullanılarak kişisel bazda öneri yapmayı hedefleyen içerik tabanlı öneri sistemi uygulamalarının geliştirilmesi bağlamında literatüre katkı sağlayacağı değerlendirilmektedir Daha fazlası Daha az

HAKAN AYDIN

Makale | 2022 | Computers & Security , pp.32 - 53

The fact that cloud systems are under the increasing risks of cyber attacks has made the phenomenon of information security first a need and then a necessity for these systems. Distributed Denial of Service (DDoS) attacks can exploit, disrupt, change, prevent or damage cloud services. Accurate and timely detection and prevention of these attacks are very important in terms of ensuring information security. During the COVID-19 period, the increase in the use of information technologies and especially the internet has made cyber attacks a real concern. Deep learning (DL) has become widely used for the purpose of detecting and preventi . . .ng cyber attacks to provide information security. In this study, a Long Short-Term Memory (LSTM) based system (LSTM-CLOUD) which was designed for the detection and prevention of DDoS attacks in a public cloud network environment was proposed. The design of the system is based on a signature-based attack detection approach. The LSTM-CLOUD has two modules defined in the study: detection and defense. The function of the first module of the system was determined as detecting the occurrence of DDoS attacks with the LSTM DL model developed in this study with an accuracy rate of 99.83% on the CICDDoS2019 data set. The function of the second module was determined as activating the defense mechanism to protect the cloud systems when attacks are detected. The comparison results showed that our LSTM model had a performance as good as those in the previous studies conducted with different DL algorithms on the same and different datasets. The results obtained show the effectiveness of the LSTM model developed in this study in detecting the occurrence of attacks. (c) 2022 Elsevier Ltd. All rights reserved Daha fazlası Daha az

HAKAN AYDIN

Makale | 2022 | Asian Pacific Journal of Cancer Prevention ( 10 ) , pp.32 - 53

Objective: To identify which Machine Learning (ML) algorithms are the most successful in predicting and diagnosing breast cancer according to accuracy rates. Methods: The “College of Wisconsin Breast Cancer Dataset”, which consists of 569 data and 30 features, was classified using Support Vector Machine (SVM), Naive Bayes (NB), Random Forest (RF), Decision Tree (DT), K-Nearest Neighbor (KNN), Logistic Regression (LR), Multilayer Perceptron (MLP), Linear Discriminant Analysis (LDA), XgBoost (XGB), Ada-Boost (ABC) and Gradient Boosting (GBC) ML algorithms. Before the classification process, the dataset was preprocessed. Sensitivity, a . . .ccuracy, and definiteness metrics were used to measure the success of the methods. Result: Compared to other ML algorithms used in the study, the GBC ML algorithm was found to be the most successful method in the classification of tumors with an accuracy of 99.12%. The XGB ML algorithm was found to be the lowest method with an accuracy rate of 88.10%. In addition, it was determined that the general accuracy rates of the 11 ML algorithms used in the study varied between 88-95%.Conclusion: When the results obtained from the ML classifiers used in the study are evaluated, the efficiency of the GBC algorithm in the classification of tumors is obvious. It can be said that the success rates obtained from 11 different ML algorithms used in the study are valuable in terms of being used to predict different cancer types Daha fazlası Daha az