Sharif Digital Repository

Cloud and distributed storage systems can provide large-scale data storage and high data reliability by adding redundancy to data. Redundant data may get lost due to the instability of distributed systems such as hardware failures. In order to maintain data availability, it is necessary to regenerate new redundant data in another node, referred to as a newcomer and this process reffered to repair process. Repair process is expected to be finished as soon as possible, because the regeneration time can influence the data reliability and availability of distributed storage systems. In this context, the general objective is to minimize the volume of actual network traffic caused by such... 

One of the most widely used protocols used on the Internet is the SSL protocol, which is also used in many applications to exchange information between the server and the user. Therefore, the analysis of this traffic can help decision makers in many analyses. In this thesis, we are going to present a mapping for feature vectors extracted from SSL traffic that will lead to improving the performance of machine learning algorithms.In this treatise, three methods for learning mapping are proposed, all of which are based on deep learning. The first method is to use a simple self-encoder for map learning that tries to learn a compact map from the input feature vector.The second method is the... 

Network traffic classification has been an essential element for security monitoring in the network security scope and also for quality of service purposes. Every now and then, new traffic classes are added to the available groups which are unknown to the system. In an security scope, the novelties are actually the zero-day attacks which can have huge effects on the system environment. There have been many methods developed for traffic classification which are able to distinguish known traffic using signatures or learning-based methods. In a real world scenario, The primary challenge that new traffic classifiers face, is to detect novelty and separate them from the previously known labels.... 

In recent years, highly recommended systems have been used in various areas. One of the approaches of these systems is a collaborative refinement that consists of three user-based, item-based, and matrix-based parsing. Matrix degradation methods are more effective because they allow us to discover the hidden features that exist between user and item interactions and help us better predict recommendations. The low-level mapping method is designed to store and process very high volume of data. In this method, after completing computations in the author’s nodes, the data is sent to the downsizing nodes, which is referred to as ”data spoofing”. It has been observed that in many applications, the... 

The aim of this research is to use machine learning methods in the design of Combinatorial Auctions. In particular, in this study, we first examine the relationship between Differential Privacy and combinatorial auctions. We propose a method based on Differential Privacy that, under certain assumptions, can transform any combinatorial auction based on machine learning methods into a Truthful auction using the Exponential Mechanism, such that all participants in the auction have no reason to misreport their Valuation Function. We also prove that in this case, using this method when the number of items is much less than the number of participants does not significantly impact the social... 

Machine learning models and deep learning, in attempting to solve complex and nonlinear problems, are not easily understandable, even for experts in these fields, due to the complexity of functions and issues involved. This lack of interpretability includes how models make decisions and their logical reasoning. Therefore, interpretability methods have gained attention in recent years. On the other hand, machine learning has entered many domains and penetrated a wide range of problems in various fields, especially in computer networks. This is crucial for internet service providers and computer network managers. Solving these problems enables the analysis of data flow structures in the... 

The complexity of machine learning models has made it difficult for end-users and even experts in the field to understand the reasoning behind the decisions made by these models. As a result, the need for explanation and interpretation of machine learning models has been increasing. One subset of machine learning models is clustering models. Despite the extensive research conducted on interpreting supervised models, very few studies have been focused on interpreting clustering models. In this research, we aim to propose algorithms for interpreting a clustering model in a model-agnostic and post-hoc manner. In this study, various methods are presented for interpreting a clustering model. The... 

The complexity and non-linearity of today’s machine learning-based systems make it difficult for both end users and experts in the field to understand the logic and reasoning behind their decisions and outputs. Explainable AI (XAI) methods have gained significant attention in real-world problems as they enhance our understanding of these models, increasing trust and improving their efficiency. By applying different explanation methods on a machine learning model, the same output is not necessarily observed, hence evaluation metrics are needed to assess and compare the quality of explanation methods based on one or more definitions of the goodness of the explanation produced by them. Several... 

Nowadays, neural networks are used as the main method in most machine learning applications. But research has shown that these models are vulnerable to adversarial attacks imperceptible changes to the input of neural networks that cause the net- work to be deceived and predict incorrectly. The importance of this issue in sensitive and security applications of neural networks, such as self-driving cars and medical diagnosis systems, becomes much higher. In recent years, many researches have been done in the field of making neural net- works robust against this threat, but in most of them, higher robustness has been provided on the basis of larger and more complex models. Few researches have... 

Almost all of the network traffic classification systems use pre-defined extracted features by the experts in computer network. These features include regular expressions, port number, information in the header of different layers and statistical feature of the flow. The main problem of the traffic analysis and anomaly detection system lies in finding appropriate features. The feature extraction is a time consuming process which needs an expert to be done. It is notable that the classification of special kinds of traffic like encrypted traffic is impossible using some subset of mentioned features.The lack of integration in feature detection and classification is also another important issue... 

As application demands for online convex optimization accelerate, the need for design-ing new methods that simultaneously cover a large class of convex functions and im-pose the lowest possible regret is highly rising. Known online optimization methods usually perform well only in specific settings, e.g., specific parameters such as the diam-eter of decision space, Lipschitz constant, and strong convexity coefficient, where their performance depends highly on the geometry of the decision space and cost functions. However, in practice, the lack of such geometric information leads to confusion in using the appropriate algorithm. To address these issues, some adaptive methods have been proposed... 

In network science, link prediction is one of the essential tasks that has been neglected. One important application of link prediction in telecommunication networks is analyzing the user's consumption pattern to provide better service. This project aims to predict future links with applications to call data using the users' call history. In previous research, there are two main approaches: 1) heuristic-based approach, and 2) deep-learning-based approach, such as graph neural networks. These methods are mainly used for processing static graphs, and therefore, we cannot generalize them to dynamic graphs. But there are many graphs which are dynamic in nature. For instance, call data records... 

By increasing the use of Internet and sharing information in this platform, servers are equipped with more robust hardware and a wider bandwidth network. But the growth of data and services is such that a server, no matter how powerful, does not have the ability to respond all users! Several servers are used to solve this problem. Data centers, content delivery networks, and so on are emerged depending on types of service. One of the important issues in these systems is how to distribute loads between servers in a proportional manner. In this thesis, we first examine the relationship between cost and response time on user requests. Cost can be considered as a communication cost or financial... 

Today, we live in an information age where all people can access the vast amount of data in the world by connecting to the Internet.Since the Internet has expanded significantly to share information, some individuals and organizations seek to be able to prevent the possible sabotage of some people by monitoring network users. Analysis of computer network traffic is one of the importance issues that many activities have been done in this area. One of the most important questions in traffic analysis is to identify the main content of traffic on the encrypted network. Numerous studies have shown that the traffic of websites visited through the Tor network, including Specific information that...