Thesis defense

thesis defense

Burak Çelik : OPTIMIZATION OF ADVANCED ENCRYPTION STANDARD (AES) ON CUDA

Msc. Candidate: Burak Çelik

Program: Cyber Security

Date: 04.09.2019 14:00 p.m

Place: A-212

Abstract: This thesis presents several optimization techniques of AES implementations on CUDA. 6 different CUDA kernels are implemented for AES-128 exhaustive search with different software designs and they are compared with each other using Nsight experiment results. Outcome of these results are used for finding the best CUDA implementation and from it, AES-128, AES-192 and AES-256 versions are created for exhaustive search, on the fly CTR and file encryption. They are compared with CPU implementations in order to decide whether GPU or CPU is the fastest considering these topics. For this comparison, two different type of CPU implementations are created which are AES-NI, using new instruction set of Intel, and basic C++. 1, 2, 4 and 8 threads versions of these implementations are compared with CUDA and results are shared. According to them, CUDA is 21, 19 and 18 times faster than the best CPU implementations for exhaustive search with respect to key length. These ratios are 4 times for CTR implementations in which 37.52 GBs of data can be encrypted each second while using CUDA. File encryption for CUDA is 22, 19 and 17 times faster than the best CPU implementations. CUDA can encrypt 31.24 GBs of data per second in this regard without considering I/O operations.

Announcement Category

Nur Didem Başkurt : Omnidirectional Hyperspectral Imaging

PhD. Candidate: Nur Didem Başkurt

Program: Information Systems

Date: 28.08.2019 10:00 a.m

Place: Conference Hall 01

Abstract: We aim to integrate hyperspectral cameras with catadioptric omnidirectional imaging systems to be able to benefit from the advantages of both. Hyperspectral imaging systems provide dense spectral information about the scene being investigated by collecting contiguous data from high number of bands on the electromagnetic spectrum. However the low spatial resolution of these sensors frequently bring about the mixing problem in remote sensing applications. Several unmixing approaches are developed in order to handle the challenging mixing problem on perspective images. On the other hand, omnidirectional imaging systems provide a 360-degree field of view in a single image, while they renounce high spatial resolution. Catadioptric images introduce a radial warping due to the structure of the mirror used in the system. This warping causes a non-uniformity in the spatial resolution which makes the unmixing problem more complicated. In this context, a novel spatial-contextual unmixing algorithm is developed specialized for the large field of view hyperspectral imaging systems. The proposed algorithms are evaluated on various real-world and simulated cases. Experimental results show that the proposed approach outperforms compared methods.

Announcement Category

Bekir Öztürk, : Semi Dynamic Light Maps

M.S. Candidate: Bekir Öztürk

Program: Multimedia Informatics

Date: 04.09.2019 / 13:30

Place: A-108

Abstract: One of the biggest challenges of real-time graphics applications is to maintain high frame rates while producing realistically lit results. Many realistic lighting effects such as indirect illumination, ambient occlusion, soft shadows and caustics are either too complex to render in real-time with today`s hardware or cause significant hits to frame rates. Light mapping technique offers to precompute the lighting of the scene to speed up expensive lighting calculations at run-time. This allows rendering high quality lights from a high number of light sources even on low-end devices. The primary drawback of this technique is that scene state that is dependent on the precomputed data cannot be changed at run-time. This includes intensity, color and position of light sources as well as position and visibility state of light map illuminated objects. This property of light maps significantly decreases the interactability of applications. In this thesis, we present a method to remove some of these restrictions at the cost of additional texture memory and small CPU/GPU workload. This allows changing color and intensity properties of selected light sources at run-time while keeping the benefits of light mapping technique. It is also becomes possible to change visibility state of selected objects. Our algorithm computes the light maps separately for each light source. Regions shadowed by each selected object are also captured and stored. These maps are later combined at run-time to correctly illuminate the scene. Despite the increase in the generation time of precomputed data, the overhead of the method at run-time is low enough to make it useful in many real-time applications.

Announcement Category

Özgür Ural, : AUTOMATIC DETECTION OF CYBER SECURITY EVENTS FROM TURKISH TWITTER STREAM AND TURKISH NEWSPAPER DATA

M.S. Candidate: Özgür Ural

Program: Cyber Security

Date: 07.08.2019 / 11:00

Place: A-108

Abstract: Cybersecurity experts scan the internet and face security events that influence users, institutions, and governments. An information security analyst regularly examines sources to stay up to date on security events in her/his domain of expertise. This may lead to a heavy workload for the information analysts if they do not have proper tools for security event investigation. For example, an information analyst may want to stay aware of cybersecurity events, such as a DDoS (Distributed Denial of Service) attack on a government agency website. The earlier they detect and understand the threats, the longer time remaining to alleviate the obstacle and to investigate the event. Therefore, information security analysts need to establish and keep situational awareness active about the security events and their likely effects. However, due to the large volume of information flow, it may be difficult for security analysts and researchers to detect and analyze security events timely. There have been attempts to solve this problem both from an academic perspective and engineering purposes.

 A recent challenge in this domain is that the internet community use different languages to share information. For instance, information about security events in Turkey is mostly shared on the internet in Turkish. The present thesis investigates the automatic detection of security incidents in Turkish by processing Twitter and news media. It proposes an automatic, Turkish specific software system that can detect cybersecurity events in real time.

Announcement Category

Zafer Şengül, : Modelling the Effects of Malware Propagation on Military Operations by Using Bayesian Network Framework

M.S. Candidate: Zafer Şengül

Program: Cyber Security

Date: 07.08.2019 / 09:45

Place: A-212

Abstract: Malware are malicious programs that cause unwanted system behavior and usually result in damage to IT systems or its users. These effects can also be seen during military operations because high-tech military weapons, command, control and communication systems are also interconnected IT systems. This thesis employs conventional models that have been used for modeling the propagation of biological diseases to investigate the spread of malware in connected systems. In particular, it proposes a probabilistic learning approach, namely Bayesian Network analysis, for developing a framework for the investigation of mixed epidemic model and combat models to characterize the propagation of malware. Compared to the classical models, which have employed formula-based representations, the results of this thesis reveal more enriched representations of the superiority of one military force over the other in probabilistic terms.

Announcement Category

Habibe Cansu Demirel, : Modeling the Tumor Specific Network Rewiring by Integrating Alternative Splicing Events with Structural Interactome

M.S. Candidate: Habibe Cansu Demirel

Program: Bioinformatics

Date: 31.07.2019

Place: A-108

Abstract: Alternative splicing is a post-transcriptional regulation which is important for the diversity of the proteome and eventually the interactome. It enables the production of multiple proteins from a single gene with different structures. In a network point of view, these structural changes can introduce new interactions or cause the loss of the existing ones. The variations in this mechanism has been associated with various diseases including cancer. In this study, we reconstructed patient specific networks with tumor specific protein isoforms by integrating the protein structures and the interaction losses they bring with. For this purpose, we collected 400 breast cancer tumors and 112 normal RNA-seq data from the Cancer Genome Atlas (TCGA) and found the transcripts that show increased expression patterns in tumor cells. We mapped these transcripts to their available protein isoforms found in UniProt. Additionally, we compiled a structural human interactome from multiple sources and aligned the missing residues on isoforms with the known/predicted protein interfaces to find potential interaction losses. At the end, we constructed two interactomes for each sample; one filtered based on the lost interfaces as a result of predominant isoforms (called “terminal set”) and one filtered based on the expression. Then, we used the same terminal set with Omics Integrator to model two sets of networks based on the two patient-specific interactomes. Finally, we compared the resulting two networks and all tumor specific networks simultaneously to reveal pathway, protein-protein interaction and protein patterns that can cluster the tumors according to their similarities. The results of our analysis will contribute to the elucidation of tumor mechanisms and will help for target selection and developing therapeutic strategies.

Announcement Category

Emre Süren, : AN EFFICIENT AND NOVEL DETECTION TECHNIQUE FOR NEXT GENERATION WEB-BASED EXPLOITATION KITS

PhD. Candidate: Emre Süren

Program: Bilişim Sistemleri

Date: 06.08.2019

Place: Konferans Salonu 01

Abstract: The prevalence and non-stop evolving technical sophistication of Exploit Kits (EKs) is one of the most challenging shifts in the modern cybercrime landscape. Over the last few years, malware infection via drive-by download attacks have been orchestrated with EK infrastructures. An EK serves various types of malicious content via several threat vectors for a variety of criminal attempts, which are mostly monetary-centric. In this dissertation, an in-depth discussion of the EK philosophy and internals is provided. A content analysis is introduced for the EK families where special contextaware properties are identified. A key observation is that while the webpage contents have drastic differences between distinct intrusions executed through the same EK, the patterns in URL addresses stay similar. This is due to the fact that auto-generated URLs by EK platforms follow specific templates. This dissertation proposes a new lightweight technique to quickly categorize unknown EK families with high accuracy leveraging machine learning algorithms with novel URL features. Rather than analyzing each URL individually, the proposed overall URL patterns approach examines all URLs associated with an EK infection. The method has been evaluated with a popular and publicly available dataset that contains 240 different real-world infection cases involving over 2250 URLs, the incidents being linked with the 4 major EK flavors that occurred throughout the year 2016. In the experiments, the system achieves up to 93.7 % clustering accuracy and up to 100 % classification accuracy with the estimators experimented.

Announcement Category

Amin Zabardast, A deep learning approach to surface reconstruction for surgical navigation during laparoscopic, endoscopic or robotic surgery

M.S. Candidate: Amin Zabardast

Program: Medical Informatics

Date: 07.08.2019

Place: A-108

Abstract: Minimally invasive surgical procedures utilize technology to provide surgeons with more functionality as well as a better perspective to help them succeed in their tasks and reduce operations risks. Surgeons usually rely on screens and cameras during minimally invasive surgeries such as Laparoscopic, Endoscopic, or Robotic Surgeries. Currently, operating rooms use information from different modalities such as Computer-Aided Tomography and Magnetic Resonance Imaging. However, the information is not integrated, and the task of extracting and combining features falls under the surgeon’s expertise. Conventional cameras, although very helpful, are not capable of transmitting every aspect of the scene including depth perception. Recently stereo cameras are being introduced to operating rooms. Utilizing stereo endoscopic equipment alongside algorithms to process the information can enable depth perception.The process of extracting depth information from stereo cameras, also known as Stereo Correspondence, is still an active research field in computer science. Understanding depth information from the view is a necessary step for reconstruction of the scene in a 3D environment. Ultimately, this reconstructed environment acts as a basis to build an Augmented Reality with extra information baked into the scene to help the surgeon. Artificial Neural Networks (ANNs), specially Convolutional Neural Networks (CNNs), have revolutionized the computer vision research in the past few years. One of the problems that researchers tried to solve using ANNs was Stereo Correspondence. There are variations of CNNs with excellent accuracy in Stereo Correspondence problem. This thesis aims to achieve surface reconstruction from in vitro stereo images of organs using Deep Neural Networks and in silico simulations.

Announcement Category

Thesis defense - Melih Öder

Title: MINING EYETRACKING DATA TO CHARACTERISE USERS AND THEIR PATTERNS OF USE 

PhD Candidate: Melih Öder

Program: Informations Systems Department 

Date: 26 June 10:00

Place: A-212

Abstract: Eye tracking studies typically collect an enormous amount of data that encodes a lot of information about the users’ behavior and characteristics on the web. However, there are not many studies that mine such data to learn and discover user characteristics and profiles. The main goal of this study is to mine eye tracking data by machine learning methods to create data models which characterise users and predict their characteristics, in particular, familiarity and gender. Detecting users’ characteristics can be used in creating adaptive user interfaces to improve user experience and interaction efficiency. In a typical eye tracking study, collected demographics data have participants’ educational backgrounds, gender, age, and familiarity degree to a web page (subjectively). In this thesis, a model focusing on the users’ familiarity degree and gender is first created based on an existing eye-tracking dataset, and then a new eye-tracking study is conducted to validate this model. The main contribution of this thesis is a machine learning approach that can be used to characterise users, in particular, familiarity and gender, based on eye-tracking data and also a tool that can be used to extract features and metrics from an eye-tracking dataset.

Announcement Category

Thesis defense - Fatma Ferda Özdemir

Title: SECURITY VISUALIZATION INFRASTRUCTURES, TECHNIQUES, AND METHODOLOGIES FOR IMPROVED ENTERPRISE SECURITY 

PhD Candidate: Fatma Ferda Özdemir

Program: Informations Systems Department 

Date: 17 June 15:30

Place: Conference Hall-01

Abstract: This thesis represents research focuses on providing designs to allow monitoring of the security status of enterprises at the organization level. The audience of this research is all enterprise level IT and security experts, and the other users who may be engaged in the security visualization designs, including the top level management. Numerous tools and programs are being used to analyze to overcome security vulnerabilities of the organizations. However, the outputs of these programs are rarely understood clearly. During the research, existing security visualization requirements and designs along with the corresponding technologies used for security visualization are examined. For the sake of being user-centric, a visualization requirements survey is held. The results of the literature review and the survey are converted to a substantial requirement set for a generic enterprise security visualization infrastructure. This infrastructure is implemented using industry best standards and the contemporary big data solutions. The resulting design is validated through the use of expert reviews. Later, one of the favorite security visualization subjects for the enterprises, namely web application security is depicted. A dashboard type holistic design to visualize black-box vulnerability test results is proposed along with forty plus metrics and measures. SIEM systems are also examined for their custom data visualization capabilities in parallel to this part of the study. Finally, security management related issues for the organizations was focused. In this part of this study , a decision support system for the optimization of security costs which relies on analytical methods and uses treemap type visualizations to visualize the threats, risks, corresponding precautions, and the costs is proposed. A real-world case study is used to demonstrate the benefits of this system.

Announcement Category

Pages

Subscribe to RSS - Thesis defense