Search

The aim of the paper is to reveal false names of accounts as a result of hackers attacks. The probability a given account to be either false or actual is calculated using methods for machine learning analysis. The suspected account will be used as a pattern and by classification techniques clusters will be formed with a respective probability this name to be false. The purpose of investigation is to determine if there exists a trend, which arises during the creation of new accounts, to detect the false ones and to discriminate them from the real ones independently of that if two types of accounts are generated with the same speed. These security systems are applied in different areas, where the security of the data in users' accounts is required. For example, they can be used in on-line voting for balloting, in studying the social opinion by inquiries, in protection of the information in different user accounts of given system etc.

In the present paper the values of the active resistance R b and the inductance L b of the electromagnetic ballast are determined analytically and by simulations with program OrCAD Capture. The characteristics U l = U l (t) of the studying discharge lamps in considered electrical circuit are measured experimentidly and then data are approximated and put in the OrCAD Capture models. The optimal results for R b and I l guarantee the RMS value of the current I l passed through each of the considered lamps. Different types of fluorescent and high pressure sodium lamps with nonlinear volt-amper characteristics are analyzed. The simulations' results are compared with the analytical ones, calculated by the measured data for U l = U l (t). Essentially they are closely to the actual ones, because the approximation of the characteristic U l = U l (t) for each studying lamp is more accurate because of the fact that it uses bigger number of basic points than the existing approximation.

This paper deals with chaotic cryptosystem and observer-based chaotic synchronization of Lozi chaotic generator for secure information transmission. Increasing the size of the secret key has been achieved thanks to the G-interval parameter method for observability analysis of a Lozi chaotic generator.

Converting symbolic sequences into complex genomic signals reveals surprising regularities of genomes, both locally and at a global scale. This approach allows using signal processing methods for the handling and analysis of nucleotide sequences, specifically for the prediction of nucleotides when knowing the preceding ones in the sequence. In this paper we propose both Feature Extraction (FE) and Principal Component Analysis (PCA) as methods to efficiently extract the major features of a genomic signal, using then a neural network to predict the next sample in the sequence.