arXiv subjects

Elena Hadzieva

Publications and source records attributed to Elena Hadzieva.

4 recordsLinked to original sources

arXiv2026

Parallel Gauss-Jordan Elimination and System Reduction for Efficient Circuit Simulation

For the purposes of electric circuit simulation, we consider an iterative simulation model based on solving systems of linear equations by Gauss-Jordan elimination (GJE) for individual moments in time. To accelerate the simulation, we propose two independent novel approaches: a parallel GJE algorithm and partial system reduction prior to the start of iterations. The former is based on a well-known strategy applied for the first time in this context, whereas the latter, to the best of our knowledge, proposes an entirely new system reduction approach. To evaluate performance, we implement these algorithms in C++ using OpenMP and run them on various input matrices. Our analyses of the individual methods show improved performance, whilst combining them maintains parallel efficiency after partial reduction on medium-sized matrices and even improves efficiency on the largest matrices on the tested machine.

cs.DC↗

arXiv2023

Matrix Profile based Anomaly Detection in Streaming Gait Data for Fall Prevention

The automatic detection of gait anomalies can lead to systems that can be used for fall detection and prevention. In this paper, we present a gait anomaly detection system based on the Matrix Profile (MP) algorithm. The MP algorithm is exact, parameter free, simple and efficient, making it a perfect candidate for on the edge deployment. We propose a gait anomaly detection system that is able to adapt to an individual's gait pattern and successfully detect anomalous steps with short latency. To evaluate the system we record a small database of enacted anomalous steps. The results show the system outperforms a more complex Neural Network baseline.

eess.SP↗

arXiv2015

Theoretical Aspects of a Design Method for Programmable NMR Voters

Almost all dependable systems use some form of redundancy in order to increase fault-tolerance. Very popular are the $N$-Modular Redundant (NMR) systems in which a majority voter chooses the voting output. However, elaborate systems require fault-tolerant voters which further give additional information besides the voting output, e.g., how many module outputs agree. Dynamically defining which set of inputs should be considered for voting is also crucial. Earlier we showed a practical implementation of programmable NMR voters that self-report the voting outcome and do self-checks. Our voter design method uses a binary matrix with specific properties that enable easy scaling of the design regarding the number of voter inputs N. Thus, an automated construction of NMR systems is possible, given the basic module and arbitrary redundancy $N$. In this paper we present the mathematical aspects of the method, i.e., we analyze the properties of the matrix that characterizes the method. We give the characteristic polynomials of the properly and erroneously built matrices in their explicit forms. We further give their eigenvalues and corresponding eigenvectors, which reveal a lot of useful information about the system. At the end, we give relations between the voter outputs and eigenpairs.

cs.DC↗

arXiv2015

Real-time Tool for Affine Transformations of Two Dimensional IFS Fractals

This work introduces a novel tool for interactive, real-time transformations of two dimensional IFS fractals. We assign barycentric coordinates (relative to an arbitrary affine basis of $\mathbb{R}^2$) to the points that constitute the image of a fractal. The tool uses some of the nice properties of the barycentric coordinates, enabling any affine transformation of the basis, done by click-and-drag, to be immediately followed by the same affine transformation of the IFS fractal attractor. In order to have a better control over the fractal, as affine basis we use a kind of minimal simplex that contains the attractor. We give theoretical grounds of the tool and then the software application.

cs.GR↗