This paper focuses on training machine learning models using the XGBoost and extremely randomized trees algorithms two datasets obtained static dynamic analysis of real malicious benign samples. We then compare their success rates—both mutually with other algorithms, such as random forest, decision tree, support vector machine, naïve Bayes which we compared in our previous work same datasets. The best performing classification models, algorithm, achieved 91.9% detection accuracy 98.2%...

We present a toolchain for computational research consisting of Sacred and two supporting tools. is an open source Python framework which aims to provide basic infrastructure running experiments independent the methods libraries used. Instead, it focuses on solving universal everyday problems, such as managing configurations, reproducing results, bookkeeping. Moreover, provides extensible basis other tools, we here: Labwatch helps with tuning hyperparameters, Sacredboard web-dashboard...

This paper explores the problem of processing immense volume measurement data arising during network traffic monitoring. Due to ever-increasing demands current networks, observing accurate information about every single flow is virtually infeasible. In many cases existing methods for reduction records are still not sufficient enough. Since knowledge flows termed as "heavy-hitters" suffices fulfill most monitoring purposes, we decided aggregate pertaining non-heavy-hitters. However, due...

Computer system security is the fastest developing segment in information technology. The conventional approach to mostly aimed at protecting system, while current trends are focusing on more aggressive forms of protection against potential attackers and intruders. One also application advanced technology based principle baits - honeypots. Honeypots specialized devices slowing down or diverting attention from critical resources allow future examination methods tools used by attackers....

This paper deals with accelerating computer vision computations using a specialized multicore architecture.Computer is one of the fastest-evolving segments science.Even though uses time-consuming methods, processing can be accelerated processor architectures.Singlecore processors are legacy, since they have reached their physical limits.The way to go use architectures, which also used accelerate in areas such as vision.This describes architecture that calculations field vision.The proposed...

In order to keep up with the massive growing and expanding development of electronics informatics fields study, education system should be able react on currently used technologies available market include them into process. Meeting new trends becomes crucial when producing people broad view, experience insight. Nowadays, Arduino based are gaining huge popularity among communities interested in electronic hardware solutions, because its availability market, relatively easy use last but not...

Information security is one of the priorities in context communication technologies. The work focused on intrusion detection high-speed computer networks. Request-transition diagrams method describing time-dependent phenomena provides an effective to announce passages preceding existence disturbance. Modified hybrid reduces responsiveness system and real-time tracking. In a correct state used only detection, which provided real-time. Definition indefinite field brings extension dual methods...

In the last twenty years there has been a huge emphasis placed on road safety due to an increasing number of drivers roads. Safety is and will remain one priority, for this reason, any positive changes or technological be welcomed by governments vehicle manufacturers. People spend more hours driving it causes not only chaos but also pollutes environment. The development new smart technology therefore become important focus many companies, institutions researchers. Smart cities are demanding...

Intrusion detection is a system pointed to identification of unauthorized use, or intrusion attempt in computer network environment. IDS systems are based on behavior prediction attacker and legitimate user network. For more accurately security events, this proposal suggests use customized DIDS (distributed systems), which combines distributed monitoring using external sensors centralized data analysis. The main problem project fusion interpretation events from multiple sensor networks....

This work discusses the formalization of sentence composition and discovery semantic ambiguities natural language.It also original connection between logic area represented by predicate linear ludics, as well as, linguistic Montague grammar.Montague grammar is a tool that allows analysing sentences in their extensional intentional contexts.Predicate non-traditional actions resources where assumptions are consumed after use implication.Ludics uses proofs from to analyse strategies actors...

Cloud computing is a highly discussed topic in the world of IT. The rapid development cloud and growing interest companies it evident everywhere world. This paper handle about using services through Internet by small medium businesses based on computing. First part focused principles evaluates its advantages disadvantages. second designing an own goal this work to create system for companies. In implementation used VMware platform experimental verifying functionality.

This paper focuses on the implementation of nfstream, an open source network data analysis tool and machine learning model using TensorFlow library for HTTP attack detection. attacks are common pose a significant security threat to networked systems. In this paper, we propose learning-based approach detect aforementioned attacks, by exploiting capabilities TensorFlow. We also focused collection traffic which provides detailed flows. pre-processed transformed collected into vectors, were used...

This paper focuses on malware analysis and detection using machine learning methods.The aim of the authors was to perform static dynamic programs designed for Windows then present results as a dataset.We analysed implemented different classification methods, such decision trees, random forests, support vectors naive Bayes methods.We verified their ability distinguish malicious harmless samples evaluated success rate accuracy metrics.Then, we compared obtained by prediction over dataset...

Ransomware problem recently proves to be a quite danger threat that after its launching ciphers the user files while storing and hiding decryption key until ransom is paid by victim. This type of malware has been lucrative business for cybercriminals black hats. The ransomware spreading easily possible because lack prevention mechanisms in traditional detection-based protection, such as antivirus anti-malware. Additionally, this forms incorporate advanced encryption algorithms expand number...

Online voting is a topic that companies have been dealing with for many years. Several countries tried and are still trying to implement such system instead of classical paper election we know the most nowadays, as it brings advantages. Blockchain technology became known people mainly thanks Bitcoin cryptocurrency which was created in 2008. presented new solution almost ideal purposes. With use blockchain, security reliability online systems has also increased over time. Voting transparent....

The work is devoted to the analysis of technologies and design a responsive social network application with elements decentralization. Despite all positives that networks bring, we can't forget potential problems associated their use. Some can be minimized by using decentralization, which has effect limiting third-party control.The centralized part created Vue.js Laravel frameworks, provide basic functions. In context very important creation smart contract stored in Ethereum blockchain. By...

The ambition to create hybrid between social network and forum (blog chat) is based on the idea of Slovak community site called "Birdz". Hybrid merge functions popular website types. It tries introduce all-in-one that join best most requested features networks, forums, blogs chats. main contribution this work concept network.

In the last decade, processing of high dimensional data became inevitable task in many areas research and daily life. Feature selection (FS), as part methodology, is an important step knowledge discovery. This paper proposes nine variation two-step feature approach with filter FS employed first exhaustive search second step. The performance proposed methods comparatively analysed from stability predictive point view. As obtained results indicate choice stage has strong influence on resulting...