Σχολή Επιστημών και Τεχνολογίας της Πληροφορίας
Μόνιμο URI για αυτήν την κοινότηταhttps://pyxida.aueb.gr/handle/123456789/2
Η Σχολή Επιστημών και Τεχνολογίας της Πληροφορίας περιλαμβάνει τα Τμήματα: - Τμήμα Πληροφορικής - Τμήμα Στατιστικής
Περιήγηση
Πλοήγηση Σχολή Επιστημών και Τεχνολογίας της Πληροφορίας ανά Επιβλέπων "Athens University of Economics and Business, Department of Informatics"
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω
Τώρα δείχνει 1 - 4 από 4
- Αποτελέσματα ανά σελίδα
- Επιλογές ταξινόμησης
Τεκμήριο Cyber security in maritime industryΓεράρδος, Παύλος; Athens University of Economics and Business, Department of InformaticsThe purpose of this thesis is to examine the role of Information and Communications Technology (ICT)in today’s maritime industry with an increased interest in its security aspects.The advantages that modern technologies have provided such as the Internet, satellite communications and industrial control systems – have made the transition to an ICT-enabled industry a necessity rather than a “nice-to-have” feature. Vessels are now able to communicate with the shore for business (spare orders, bills of lading) and personal purposes (Internet access), get real-time information for safe navigation (digital charts and weather updates) as well as monitor their systems for operational and performance status (engine & cargo control). Port infrastructures are now also able to support more efficiently cargo movement operations and thus provide services to critical supply chains of local, state or national levels.Although this transition has pushed the industry forward, it also brought to the fore several new risks that all involved parties should understand and take under consideration. The reliance on digital communications and automation systems for shipping operations make cyber security a global issue with possible cascading effects, which could starton a company-level but indirectly affect security and welfare of third parties, associates or even whole states and nations. Emerging cyber threats including destructive malware, cybercrime and data leaks impact governments and industry alike.By the term “ICT” we refer to all Information Technology (IT) and Operational Technology (OT) equipment which can be found on shore and at sea, and more specifically in:*Shipping company offices (IT)*Vessels (IT, OT)*Ports (IT, OT)as well as the communication technologies which allow for inter-connectivity between them.The main focus of this thesis will be the vessels, as it is the one that is differentiating from the offices and ports due to its completely remote nature.Τεκμήριο Research on economic mechanisms and incentives for technologies of future internetKostopoulos, Alexandros G.; Courcoubetis, Costas; Athens University of Economics and Business, Department of InformaticsThis dissertation delves into various aspects about the adoption of new protocols in today‘s Internet. The deployment and adoption of Internet protocols differ from the diffusion of end user-centred innovations, such as consumer-products, since the Internet is a complex system consisting of different stakeholders with conflicting interests. It is suggested that the Internet architecture should be flexible enough to sufficiently incorporate stakeholders‘ preferences, the so-called ―design for tussle‖ principle. We analyse the proposed ―design for tussle‖ principle by investigating selected protocols of current Internet and we relate adoption issues to their compatibility with this principle. Although tussle-aware design is very important for the long-term evolution of Internet, it is not sufficient condition to ensure the adoption of a certain protocol. Apart from the tussle-aware design, we consider several factors affecting the adoption of a new protocol. Based on these factors, we propose an adoption framework for Internet protocols, which can be used as a roadmap for analysing their potential deployment and adoption by stakeholders. We concentrate on emerging protocols for the Internet intending to solve bandwidth allocation issues better; Multipath TCP (MPTCP), as well as protocols for information exposure of the network congestion. We present their basic operation, and we investigate how such mechanisms incorporate the tussles on bandwidth sharing and resource allocation into the Internet. Moreover, we apply the aforementioned adoption framework to these protocols in order to give a broader picture of the issues surrounding their adoption. We investigate the new challenges for the involved stakeholders, the viability of the different deployment scenarios, as well as the business models that may be enabled. Finally, we consider how to give the right incentives to the involved stakeholders for adopting these proposed Internet protocols.Τεκμήριο Services architecture on top of the peer-to-peer wireless network confederation(Athens University of Economics and Business, 06-2006) Paraskevaidis, Dimitris D.; Polyzos, George; Athens University of Economics and Business, Department of InformaticsThesis - Athens University of Economics and Business. Postgraduate, Department of InformaticsΤεκμήριο Statistical and computational efficiency comparison between MCMC and Hamiltonian Monte Carlo with application in epidemiological models(2018) Barmpounakis, Petros; Μπαρμπουνάκης, Πέτρος; Demiris, Nikolaos; Athens University of Economics and Business, Department of InformaticsIn the recent years several methods have been proposed and applied to different data problems inorder to perform Bayesian inference for the posterior density of the model’s parameters. Fornearly three decades the most widely used are Markov Chain Monte Carlo methods andspecifically Metropolis-Hastings and Gibbs Sampling, which will be examined for the purposesof this thesis. The new hot topic in Bayesian Inference is Hamiltonian Monte Carlo, whichpromises respectable results as regards accuracy and speed. In epidemiology it is even morecrucial to find an algorithm that returns the most accurate results in the most efficient way. Inthis project, the two aforementioned algorithms will be implemented in a temporal-stochasticepidemiology model regarding the spreading of the Sheeppox virus in the region of Evros,Greece. The two algorithms will be compared by their computational efficiency based on theCPU time and their statistical efficiency after comparing the results returned from the prequentialanalysis, using appropriate scoring rules. At this point, it is of crucial importance to highlight thedifferent softwares used to implement the algorithms; Metropolis-Hastings and Gibbs Samplingrun through WinBUGS and Hamiltonian Monte Carlo through the new probabilistic languageStan. Consequently, any differences occurring in the results may also be derivatives of thedifferent softwares usage. At this stage, it is important to mention that Stan is on experimentallevel; hence some inaccuracies in the results may occur. The interface for both softwares ischosen to be R.