Magazine et Edition

Volume 22 (2022): Edition 3 (September 2022)

Volume 22 (2022): Edition 2 (June 2022)

Volume 22 (2022): Edition 1 (March 2022)

Volume 21 (2021): Edition 4 (December 2021)

Volume 21 (2021): Edition 3 (September 2021)

Volume 21 (2021): Edition 2 (June 2021)

Volume 21 (2021): Edition 1 (March 2021)

Volume 20 (2020): Edition 6 (December 2020)
Special Edition on New Developments in Scalable Computing

Volume 20 (2020): Edition 5 (December 2020)
Special issue on Innovations in Intelligent Systems and Applications

Volume 20 (2020): Edition 4 (November 2020)

Volume 20 (2020): Edition 3 (September 2020)

Volume 20 (2020): Edition 2 (June 2020)

Volume 20 (2020): Edition 1 (March 2020)

Volume 19 (2019): Edition 4 (November 2019)

Volume 19 (2019): Edition 3 (September 2019)

Volume 19 (2019): Edition 2 (June 2019)

Volume 19 (2019): Edition 1 (March 2019)

Volume 18 (2018): Edition 5 (May 2018)
Special Thematic Edition on Optimal Codes and Related Topics

Volume 18 (2018): Edition 4 (November 2018)

Volume 18 (2018): Edition 3 (September 2018)

Volume 18 (2018): Edition 2 (June 2018)

Volume 18 (2018): Edition 1 (March 2018)

Volume 17 (2017): Edition 5 (December 2017)
Special Edition With Selected Papers From The Workshop “Two Years Avitohol: Advanced High Performance Computing Applications 2017

Volume 17 (2017): Edition 4 (November 2017)

Volume 17 (2017): Edition 3 (September 2017)

Volume 17 (2017): Edition 2 (June 2017)

Volume 17 (2017): Edition 1 (March 2017)

Volume 16 (2016): Edition 6 (December 2016)
Special issue with selection of extended papers from 6th International Conference on Logistic, Informatics and Service Science LISS’2016

Volume 16 (2016): Edition 5 (October 2016)
Edition Title: Special Edition on Application of Advanced Computing and Simulation in Information Systems

Volume 16 (2016): Edition 4 (December 2016)

Volume 16 (2016): Edition 3 (September 2016)

Volume 16 (2016): Edition 2 (June 2016)

Volume 16 (2016): Edition 1 (March 2016)

Volume 15 (2015): Edition 7 (December 2015)
Special Edition on Information Fusion

Volume 15 (2015): Edition 6 (December 2015)
Special Edition on Logistics, Informatics and Service Science

Volume 15 (2015): Edition 5 (April 2015)
Special Edition on Control in Transportation Systems

Volume 15 (2015): Edition 4 (November 2015)

Volume 15 (2015): Edition 3 (September 2015)

Volume 15 (2015): Edition 2 (June 2015)

Volume 15 (2015): Edition 1 (March 2015)

Volume 14 (2014): Edition 5 (December 2014)
Special Edition

Volume 14 (2014): Edition 4 (December 2014)

Volume 14 (2014): Edition 3 (September 2014)

Volume 14 (2014): Edition 2 (June 2014)

Volume 14 (2014): Edition 1 (March 2014)

Volume 13 (2013): Edition Special-Edition (December 2013)

Volume 13 (2013): Edition 4 (December 2013)
The publishing of the present issue (Volume 13, No 4, 2013) of the journal “Cybernetics and Information Technologies” is financially supported by FP7 project “Advanced Computing for Innovation” (ACOMIN), grant agreement 316087 of Call FP7 REGPOT-2012-2013-1.

Volume 13 (2013): Edition 3 (September 2013)

Volume 13 (2013): Edition 2 (June 2013)

Volume 13 (2013): Edition 1 (March 2013)

Volume 12 (2012): Edition 4 (December 2012)

Volume 12 (2012): Edition 3 (September 2012)

Volume 12 (2012): Edition 2 (June 2012)

Volume 12 (2012): Edition 1 (March 2012)

Détails du magazine
Format
Magazine
eISSN
1314-4081
Première publication
13 Mar 2012
Période de publication
4 fois par an
Langues
Anglais

Chercher

Volume 20 (2020): Edition 6 (December 2020)
Special Edition on New Developments in Scalable Computing

Détails du magazine
Format
Magazine
eISSN
1314-4081
Première publication
13 Mar 2012
Période de publication
4 fois par an
Langues
Anglais

Chercher

20 Articles
Accès libre

Preface

Publié en ligne: 31 Dec 2020
Pages: 3 - 4

Résumé

Abstract

We are pleased to present the special issue “New developments in scalable computing” of the scientific journal “Cybernetics and Information Technologies”. For this issue (Volume 20, No 6 – December 2020), we have selected 19 papers which have gone through peer review and represent novel results in the field of Scalable Computing using state-of-the-art high-performance computing infrastructures.

Accès libre

Performance Optimization System for Hadoop and Spark Frameworks

Publié en ligne: 31 Dec 2020
Pages: 5 - 17

Résumé

Abstract

The optimization of large-scale data sets depends on the technologies and methods used. The MapReduce model, implemented on Apache Hadoop or Spark, allows splitting large data sets into a set of blocks distributed on several machines. Data compression reduces data size and transfer time between disks and memory but requires additional processing. Therefore, finding an optimal tradeoff is a challenge, as a high compression factor may underload Input/Output but overload the processor. The paper aims to present a system enabling the selection of the compression tools and tuning the compression factor to reach the best performance in Apache Hadoop and Spark infrastructures based on simulation analyzes.

Mots clés

  • Hadoop
  • Spark
  • data compression
  • CPU/IO tradeoff
  • performance optimization
Accès libre

New Uniform Subregular Parallelisms of PG(3, 4) Invariant under an Automorphism of Order 2

Publié en ligne: 31 Dec 2020
Pages: 18 - 27

Résumé

Abstract

A spread in PG(n, q) is a set of lines which partition the point set. A parallelism is a partition of the set of lines by spreads. A parallelism is uniform if all its spreads are isomorphic. Up to isomorphism, there are three spreads of PG(3, 4) – regular, subregular and aregular. Therefore, three types of uniform parallelisms are possible. In this work, we consider uniform parallelisms of PG(3, 4) which possess an automorphism of order 2. We establish that there are no regular parallelisms, and that there are 8253 nonisomorphic subregular parallelisms. Together with the parallelisms known before this work, this yields a total of 8623 known subregular parallelisms of PG(3, 4).

Mots clés

  • Finite Geometry
  • Combinatorics
  • Classification
  • Isomorphism
  • Parallel Computing
  • Spread
  • Parallelism
  • MPI
Accès libre

Two-Stage Search-Based Approach for Determining and Sorting of Mountain Hiking Routes Using Directed Weighted Multigraph

Publié en ligne: 31 Dec 2020
Pages: 28 - 39

Résumé

Abstract

The mountain hiking destinations become more popular as this is one of the possible ways to cope with workplace stress and to prevent burnout. In contrast to the tourist destinations, mountain hiking requires special attention due to the variety of mountain trails satisfying the same starting and finishing point for a particular route. For the goal, a two-stage search-based approach for a determining of possible routes considering the users’ preferences is developed. The first stage is focused on the determining of possible hiking routes taking into account the requirements and tourists’ preferences, while the second stage concerns the sorting of already determined hiking routes. The applicability of the described approach is illustrated and the obtained results demonstrate the capability in searching and sorting of mountain hiking trails using directed weighted multigraph including tourists’ preferences.

Mots clés

  • Directed weighted multigraph
  • Search-based approach
  • Mountain hiking algorithm
  • Mountain routes sorting
  • Geographic scalability
Accès libre

New Local Search Procedure for Workforce Planning Problem

Publié en ligne: 31 Dec 2020
Pages: 40 - 48

Résumé

Abstract

Optimization of workforce planning is important for any production area. This leads to an improvement in production process. The aim is minimization of the assignment costs of the workers, who will do the jobs. The problem is to select and assign employers to the jobs to be performed. The constraints are very strong, coming both from the specifics of the production process and from the legislation. Sometimes it is difficult to find feasible solutions. The complexity of the problem is very high and the needed number of calculations is exponential, therefore only specially developed algorithms can be applied. The objective is to minimize the assignment cost, while respecting all requirements. We propose a new hybrid metaheuristic algorithm to solve the workforce-planning problem, which is a combination between Ant Colony Optimization (ACO) and suitable local search procedure.

Mots clés

  • Metaheuristics
  • ant colony optimization
  • local search
  • workforce planning
  • combinatorial optimization
Accès libre

High Performance Machine Learning Models of Large Scale Air Pollution Data in Urban Area

Publié en ligne: 31 Dec 2020
Pages: 49 - 60

Résumé

Abstract

Preserving the air quality in urban areas is crucial for the health of the population as well as for the environment. The availability of large volumes of measurement data on the concentrations of air pollutants enables their analysis and modelling to establish trends and dependencies in order to forecast and prevent future pollution. This study proposes a new approach for modelling air pollutants data using the powerful machine learning method Random Forest (RF) and Auto-Regressive Integrated Moving Average (ARIMA) methodology. Initially, a RF model of the pollutant is built and analysed in relation to the meteorological variables. This model is then corrected through subsequent modelling of its residuals using the univariate ARIMA. The approach is demonstrated for hourly data on seven air pollutants (O3, NOx, NO, NO2, CO, SO2, PM10) in the town of Dimitrovgrad, Bulgaria over 9 years and 3 months. Six meteorological and three time variables are used as predictors. High-performance models are obtained explaining the data with R2 = 90%-98%.

Mots clés

  • Machine learning
  • Random Forest
  • Autoregressive integrated moving average
  • error correction
  • time series
  • forecasting
Accès libre

Evaluating Machine Learning Approaches for Discovering Optimal Sets of Projection Operators for Quantum State Tomography of Qubit Systems

Publié en ligne: 31 Dec 2020
Pages: 61 - 73

Résumé

Abstract

Finding optimal measurement schemes in quantum state tomography is a fundamental problem in quantum computation. It is known that for non-degenerate operators the optimal measurement scheme is based on mutually unbiassed bases. This paper is a follow up from our previous work, where we use standard numerical approaches to look for optimal measurement schemes, where the measurement operators are projections on individual pure quantum states. In this paper we demonstrate the usefulness of several machine learning techniques – reinforcement learning and parallel machine learning approaches, to discover measurement schemes, which are significantly better than the ones discovered by standard numerical methods in our previous work. The high-performing quorums of projection operators we have discovered have complex structure and symmetries, which may imply that the optimal solution will possess such symmetries.

Mots clés

  • Quantum information
  • optimization problem
  • Widening
  • reinforcement learning
Accès libre

A New Class of “Growth Functions” with Polynomial Variable Transfer Generated by Real Reaction Networks

Publié en ligne: 31 Dec 2020
Pages: 74 - 81

Résumé

Abstract

In [4, 5], two classes of growth models with “exponentially variable transfer” and “correcting amendments of Bateman-Gompertz-Makeham-type” based on a specific extended reaction network have been studied [1]. In this article we will look at the new scheme with “polynomial variable transfer”. The consideration of such a dynamic model in the present article is dictated by our passionate desire to offer an adequate model with which to well approximate specific data in the field of computer viruses propagation, characterized by rapid growth in the initial time interval. Some numerical examples, using CAS Mathematica illustrating our results are given.

Mots clés

  • Reaction networks
  • Generalized growth model
  • Exponentially and polynomial variable transfers
Accès libre

Comments on the Half Logistic Inverse Rayleigh (HLIR) cdf with “Polynomial Variable Transfer”. Some Applications

Publié en ligne: 31 Dec 2020
Pages: 82 - 93

Résumé

Abstract

On the base of the Half Logistic – G family of distributions proposed by Cordeiro, Alizadeh and Marinho [2] some mathematical properties are investigated by Almarashi et al. [1]. We study the “saturation” to the horizontal asymptote: t=1 by the new growth function M(t) in the Hausdorff sense. Similar to our previous studies [3-6], in this article we will define and analyze in detail the new family. We will call this family the “Half-Logistic-Inverse-Rayleigh cdf with Polynomial Variable Transfer” (HLIRPVT) cdf. Section 3 shows the potentiality of proposed new model under four real data sets. Some numerical examples using CAS MATHEMATICA are given.

Mots clés

  • Half-Logistic-Inverse-Rayleigh (HLIR) cdf
  • Half-Logistic-Inverse-Rayleigh cdf with Polynomial Variable Transfer (HLIRPVT) cdf
  • “Saturation” in Hausdorff sense
  • Hausdorff distance
  • Upper and lower bounds
Accès libre

Performance Analysis of a Scalable Algorithm for 3D Linear Transforms on Supercomputer with Intel Processors/Co-Processors

Publié en ligne: 31 Dec 2020
Pages: 94 - 104

Résumé

Abstract

Practical realizations of 3D forward/inverse separable discrete transforms, such as Fourier transform, cosine/sine transform, etc. are frequently the principal limiters that prevent many practical applications from scaling to a large number of processors. Existing approaches, which are based primarily on 1D or 2D data decompositions, prevent the 3D transforms from effectively scaling to the maximum (possible/available) number of computer nodes. A highly scalable approach to realize forward/inverse 3D transforms has been proposed. It is based on a 3D decomposition of data and geared towards a torus network of computer nodes. The proposed algorithms requires compute-and-roll time-steps, where each step consists of an execution of multiple GEMM operations and concurrent movement of cubical data blocks between nearest neighbors. The aim of this paper is to present an experimental performance study of an implementation on high performance computer architecture.

Mots clés

  • 3D linear transforms
  • parallel implementation
  • Intel processors/ co-processors
Accès libre

Numerical Solution of Integro-Differential Equations Modelling the Dynamic Behavior of a Nano-Cracked Viscoelastic Half-Plane

Publié en ligne: 31 Dec 2020
Pages: 105 - 115

Résumé

Abstract

The scattering of time-harmonic waves by a finite, blunt nano-crack in a graded, viscoelastic bulk material with a free surface is considered in this work. Non-classical boundary conditions and a localized constitutive equation at the interface between crack and matrix, following the Gurtin-Murdoch surface elasticity theory are introduced. An efficient numerical technique is developed using integro-differential equations along the nano-crack line that is based on an analytically derived Green‘s function for the quadratically inhomogeneous half-plane. The dependence of the diffracted and scattered waves and of the local stress concentration fields on key problem parameters such as viscosity, inhomogeneity, surface elasticity, and interaction between the nano-crack and the free surface are all examined through an extensive parametric study.

Mots clés

  • Viscoelasticity
  • graded half-plane
  • nano-crack
  • integro-differential equations
  • stress concentration
  • wave scattering
  • wave diffraction
Accès libre

On Dynamic Parallelization of Multilevel Monte Carlo Algorithm

Publié en ligne: 31 Dec 2020
Pages: 116 - 125

Résumé

Abstract

MultiLevel Monte Carlo (MLMC) attracts great interest for numerical simulations of Stochastic Partial Differential Equations (SPDEs), due to its superiority over the standard Monte Carlo (MC) approach. MLMC combines in a proper manner many cheap fast simulations with few slow and expensive ones, the variance is reduced, and a significant speed up is achieved. Simulations with MC/MLMC consist of three main components: generating random fields, solving deterministic problem and reduction of the variance. Each part is subject to a different degree of parallelism. Compared to the classical MC, MLMC introduces “levels” on which the sampling is done. These levels have different computational cost, thus, efficiently utilizing the parallel resources becomes a non-trivial problem. The main focus of this paper is the parallelization of the MLMC Algorithm.

Mots clés

  • SPDE
  • MLMC
  • UQ
  • Parallelization
  • Flow in random porous media
Accès libre

Geometrically Non-Linear Vibration of a Cantilever Interacting with Rarefied Gas Flow

Publié en ligne: 31 Dec 2020
Pages: 126 - 139

Résumé

Abstract

The work is devoted to study 2D pressure driven rarefied gas flow in a microchannel having an elastic obstacle. The elastic obstacle is clamped at the bottom channel wall and its length is half of the channel height. The gas flow is simulated by Direct Simulation Monte Carlo (DSMC) method applying the advanced Simplified Bernoulli Trial (SBT) collision scheme. The elastic obstacle is modelled as geometrically nonlinear Euler Bernoulli beam. A reduced 3 modes reduction model of the beam is created. The influence of the gas flow on the beam vibration is studied, considering the linear and nonlinear beam theories.

Mots clés

  • Fluid-Structure Interaction (FSI)
  • rarefied gas
  • microchannel
  • DSMC
  • Euler-Bernoulli beam
  • pressure driven flow
Accès libre

Molecular Dynamics Simulations of Acetylcholinesterase – Beta-Amyloid Peptide Complex

Publié en ligne: 31 Dec 2020
Pages: 140 - 154

Résumé

Abstract

Alzheimer’s Disease (AD) is a neurodegenerative disorder with severe consequences and lethal outcome. One of the pathological hallmarks of the disease is the formation of insoluble intercellular beta-Amyloid (Aβ) plaques. The enzyme ACetylcholinEsterase (AChE) promotes and accelerates the aggregation of toxic Aβ protofibrils progressively converted into plaques. The Peripheral Anionic Site (PAS), part of the binding gorge of AChE, is one of the nucleation centers implicated in the Aβ aggregation. In this study, the Aβ peptide was docked into the PAS and the stability of the formed complex was investigated by molecular dynamics simulation for 1 μs (1000 ns). The complex was stable during the simulation. Apart from PAS, the Aβ peptide makes several additional contacts with AChE. The main residence area of Aβ on the surface of AChE is the region 344-361. This region is next to PAS but far enough to be sterically hindered by dual-site binding AChE inhibitors.

Mots clés

  • Alzheimer’s disease
  • beta-Amyloid (Aβ)
  • acetylcholinesterase (AChE)
  • Peripheral Anionic Site (PAS)
  • Aβ aggregation
  • senile plaques
  • amyloid fibrils
  • neurodegenerative disorder
  • molecular dynamics
Accès libre

Trend Analysis of CMIP5 Ensemble of Climate Indices over Southeast Europe with Focus on Agricultural Impacts

Publié en ligne: 31 Dec 2020
Pages: 155 - 165

Résumé

Abstract

Nowadays there is a strong degree of agreement that the climate change is the defining challenge of our time, which will exert influence on the ecosystems, on all branches of the international economy and on the quality of life. The analysis based on climate indices is widely used non-parametric approach for quantification of the mean state as well as extreme climate events. This study, which is continuation of our previous efforts, is dedicated to the assessment of the trend magnitude and the trend statistical significance of six temperature-based and three precipitation-based indices in projected future climate over Southeast Europe up to the end of the 21st century. The indices are computed from the bias-corrected output of five CMIP5 global models, reinforced with all four RCP emission scenarios. The model output is accessed from the section of the Inter Sectoral Impact Model Intercomparison Project in the Copernicus Data Store. The multi model ensemble medians of the temperature-based indices shows considerable increase which is consistent with the warming of the mean temperatures. These changes are statistically significant in most cases and intensify with the radiative forcing. The revealed tendencies of the precipitation-based indices are more complex when compared with temperature tendencies.

Mots clés

  • Climate indices
  • RCP
  • CMIP5 Ensemble
  • future climate
  • trend analysis
Accès libre

Degree-Day Climatology over Central and Southeast Europe for the Period 1961-2018 – Evaluation in High Resolution

Publié en ligne: 31 Dec 2020
Pages: 166 - 174

Résumé

Abstract

The ongoing climate change over Central and Southeast Europe has a great potential to affect significantly the public energy demands and in particular the energy consumption in the residential heating and cooling sector. The linkage of the ambient daily extreme and mean temperatures and the energy needs for condition or heat buildings can be quantified as numerical indicators as the heating and cooling degree-days. In the present study, these indicators are calculated according the UK Met Office methodology from the daily mean and extreme temperatures, which, in turn, are computed from the output of the MESCAN-SURFEX system in the frame the FP7 UERRA project. The study, which is performed in a very high resolution, is dedicated on the analysis of the spatial patterns as well as assessment of the magnitude and statistical significance of the temporal evolution of the heating and cooling degree-days. It reveals general tendencies which are coherent with the regional climate warming, but with high spatial heterogeneities. The study confirms the essential impact of the ongoing climate change on the heating, ventilating and air-conditioning industry over Central and Southeast Europe.

Mots clés

  • Degree-days
  • MESCAN-SURFEX
  • UERRA Project
  • trend analysis
Accès libre

AllerScreener – A Server for Allergenicity and Cross-Reactivity Prediction

Publié en ligne: 31 Dec 2020
Pages: 175 - 184

Résumé

Abstract

Allergenicity of proteins is a subtle property encoded in their structures. The prediction of allergenicity of novel proteins saves time and resources for subsequent experimental work. In the host antigen-presenting cells, the allergens are processed as antigens by the means of Human Leukocyte Antigens (HLA) class II proteins. Sometimes, people allergic to a given protein show allergic reaction to a different protein, even when the two proteins have different routes of exposure. This phenomenon is termed cross-reactivity. Here, we describe a server for allergenicity and cross-reactivity prediction based on the abilities of allergenic proteins to generate binders to HLA class II proteins. The generated peptides are compared to HLA binders originating from known allergens. As a result, the server returns a list of common binders, origin proteins, and species. Different species generate common HLA binders and this determines their cross-reactivity. The server is named AllerScreener and is freely accessible at: http://www.ddg-pharmfac.net/AllerScreener.

Mots clés

  • AllerScreener
  • allergenicity
  • cross-reactivity
  • EpiTOP
  • EpiDOCK
  • HLA class II
Accès libre

Design of Multi-Epitope Vaccine against SARS-CoV-2

Publié en ligne: 31 Dec 2020
Pages: 185 - 193

Résumé

Abstract

The ongoing COVID-19 pandemic requires urgently specific therapeutics and approved vaccines. Here, the four structural proteins of the Severe Acute Respiratory Syndrome CoronaVirus 2 (SARS-CoV-2), the causative agent of COVID-19, are screened by in-house immunoinformatic tools to identify peptides acting as potential T-cell epitopes. In order to act as an epitope, the peptide should be processed in the host cell and presented on the cell surface in a complex with the Human Leukocyte Antigen (HLA). The aim of the study is to predict the binding affinities of all peptides originating from the structural proteins of SARS-CoV-2 to 30 most frequent in the human population HLA proteins of class I and class II and to select the high binders (IC50 < 50 nM). The predicted high binders are compared to known high binders from SARS-CoV conserved in CoV-2 and 77% of them coincided. The high binders will be uploaded onto lipid nanoparticles and the multi-epitope vaccine prototype will be tested for ability to provoke T-cell mediated immunity and protection against SARS-CoV-2.

Mots clés

  • COVID-19
  • multi-epitope vaccine
  • SARS-CoV
  • SARS-CoV-2
  • EpiJen
  • EpiTOP
  • EpiDOCK
  • high binders
  • HLA class I
  • HLA class II
Accès libre

Immunoinformatic Analysis of Human Thyroglobulin

Publié en ligne: 31 Dec 2020
Pages: 194 - 200

Résumé

Abstract

The AutoImmune ThyroiDitis (AITD), known as Hashimoto’s disease, is a chronic autoimmune thyroid disease progressively developed to hypothyroidism. The AITD is characterized by the formation of autoantibodies targeting two specific thyroid antigens, Thyroglobulin (Tg) and Thyroid PerOxidase (TPO). Tg is a precursor of the thyroid hormones while TPO catalyses their synthesis. The AITD has a strong genetic predisposition. During the last years, it was found that the susceptibility to AITD is associated with certain Human Leukocyte Antigens (HLA) class II genes of loci DR and DQ. In the present study, we applied in-house immunoinformatic tools to identify peptides originating from Tg and binding to AITD susceptible alleles: HLA-DR3, HLA-DR4, HLA-DR5, HLA-DQ2 and HLA-DQ8. Five peptide fragments containing promiscuous overlapping binders were selected. These were p470, p949, p1948, p2348 and p2583. Only one of them contains a known epitope (p1948). The rest have not been reported yet. The selected peptide fragments will be coupled to monoclonal antibodies specific to inhibitory B cell receptors designed to suppress the production of Tg autoantibodies.

Accès libre

Finite-Temperature Single Molecule Vibrational Dynamics from Combined Density Functional Tight Binding Extended Lagrangian Dynamics Simulations and Time Series Analysis

Publié en ligne: 31 Dec 2020
Pages: 201 - 212

Résumé

Abstract

Combining a computationally efficient and affordable molecular dynamics approach, based on atom-centered density matrix propagation scheme, with the density functional tight binding semiempirical quantum mechanics, we study the vibrational dynamics of a single molecule at series of finite temperatures, spanning quite wide range. Data generated by molecular dynamics simulations are further analyzed and processed using time series analytic methods, based on correlation functions formalism, leading to both vibrational density of states spectra and infrared absorption spectra at finite temperatures. The temperature-induced dynamics in structural intramolecular parameters is correlated to the observed changes in the spectral regions relevant to molecular detection. In particular, we consider a case when an intramolecular X-H stretching vibrational states are notably dependent on the intramolecular torsional degree of freedom, the dynamics of which is, on the other hand, strongly temperature-dependent.

Mots clés

  • Atom-centered density matrix propagation
  • density functional tight binding
  • molecular dynamics
  • finite-temperature vibrational dynamics
  • formic acid
  • torsional motion
20 Articles
Accès libre

Preface

Publié en ligne: 31 Dec 2020
Pages: 3 - 4

Résumé

Abstract

We are pleased to present the special issue “New developments in scalable computing” of the scientific journal “Cybernetics and Information Technologies”. For this issue (Volume 20, No 6 – December 2020), we have selected 19 papers which have gone through peer review and represent novel results in the field of Scalable Computing using state-of-the-art high-performance computing infrastructures.

Accès libre

Performance Optimization System for Hadoop and Spark Frameworks

Publié en ligne: 31 Dec 2020
Pages: 5 - 17

Résumé

Abstract

The optimization of large-scale data sets depends on the technologies and methods used. The MapReduce model, implemented on Apache Hadoop or Spark, allows splitting large data sets into a set of blocks distributed on several machines. Data compression reduces data size and transfer time between disks and memory but requires additional processing. Therefore, finding an optimal tradeoff is a challenge, as a high compression factor may underload Input/Output but overload the processor. The paper aims to present a system enabling the selection of the compression tools and tuning the compression factor to reach the best performance in Apache Hadoop and Spark infrastructures based on simulation analyzes.

Mots clés

  • Hadoop
  • Spark
  • data compression
  • CPU/IO tradeoff
  • performance optimization
Accès libre

New Uniform Subregular Parallelisms of PG(3, 4) Invariant under an Automorphism of Order 2

Publié en ligne: 31 Dec 2020
Pages: 18 - 27

Résumé

Abstract

A spread in PG(n, q) is a set of lines which partition the point set. A parallelism is a partition of the set of lines by spreads. A parallelism is uniform if all its spreads are isomorphic. Up to isomorphism, there are three spreads of PG(3, 4) – regular, subregular and aregular. Therefore, three types of uniform parallelisms are possible. In this work, we consider uniform parallelisms of PG(3, 4) which possess an automorphism of order 2. We establish that there are no regular parallelisms, and that there are 8253 nonisomorphic subregular parallelisms. Together with the parallelisms known before this work, this yields a total of 8623 known subregular parallelisms of PG(3, 4).

Mots clés

  • Finite Geometry
  • Combinatorics
  • Classification
  • Isomorphism
  • Parallel Computing
  • Spread
  • Parallelism
  • MPI
Accès libre

Two-Stage Search-Based Approach for Determining and Sorting of Mountain Hiking Routes Using Directed Weighted Multigraph

Publié en ligne: 31 Dec 2020
Pages: 28 - 39

Résumé

Abstract

The mountain hiking destinations become more popular as this is one of the possible ways to cope with workplace stress and to prevent burnout. In contrast to the tourist destinations, mountain hiking requires special attention due to the variety of mountain trails satisfying the same starting and finishing point for a particular route. For the goal, a two-stage search-based approach for a determining of possible routes considering the users’ preferences is developed. The first stage is focused on the determining of possible hiking routes taking into account the requirements and tourists’ preferences, while the second stage concerns the sorting of already determined hiking routes. The applicability of the described approach is illustrated and the obtained results demonstrate the capability in searching and sorting of mountain hiking trails using directed weighted multigraph including tourists’ preferences.

Mots clés

  • Directed weighted multigraph
  • Search-based approach
  • Mountain hiking algorithm
  • Mountain routes sorting
  • Geographic scalability
Accès libre

New Local Search Procedure for Workforce Planning Problem

Publié en ligne: 31 Dec 2020
Pages: 40 - 48

Résumé

Abstract

Optimization of workforce planning is important for any production area. This leads to an improvement in production process. The aim is minimization of the assignment costs of the workers, who will do the jobs. The problem is to select and assign employers to the jobs to be performed. The constraints are very strong, coming both from the specifics of the production process and from the legislation. Sometimes it is difficult to find feasible solutions. The complexity of the problem is very high and the needed number of calculations is exponential, therefore only specially developed algorithms can be applied. The objective is to minimize the assignment cost, while respecting all requirements. We propose a new hybrid metaheuristic algorithm to solve the workforce-planning problem, which is a combination between Ant Colony Optimization (ACO) and suitable local search procedure.

Mots clés

  • Metaheuristics
  • ant colony optimization
  • local search
  • workforce planning
  • combinatorial optimization
Accès libre

High Performance Machine Learning Models of Large Scale Air Pollution Data in Urban Area

Publié en ligne: 31 Dec 2020
Pages: 49 - 60

Résumé

Abstract

Preserving the air quality in urban areas is crucial for the health of the population as well as for the environment. The availability of large volumes of measurement data on the concentrations of air pollutants enables their analysis and modelling to establish trends and dependencies in order to forecast and prevent future pollution. This study proposes a new approach for modelling air pollutants data using the powerful machine learning method Random Forest (RF) and Auto-Regressive Integrated Moving Average (ARIMA) methodology. Initially, a RF model of the pollutant is built and analysed in relation to the meteorological variables. This model is then corrected through subsequent modelling of its residuals using the univariate ARIMA. The approach is demonstrated for hourly data on seven air pollutants (O3, NOx, NO, NO2, CO, SO2, PM10) in the town of Dimitrovgrad, Bulgaria over 9 years and 3 months. Six meteorological and three time variables are used as predictors. High-performance models are obtained explaining the data with R2 = 90%-98%.

Mots clés

  • Machine learning
  • Random Forest
  • Autoregressive integrated moving average
  • error correction
  • time series
  • forecasting
Accès libre

Evaluating Machine Learning Approaches for Discovering Optimal Sets of Projection Operators for Quantum State Tomography of Qubit Systems

Publié en ligne: 31 Dec 2020
Pages: 61 - 73

Résumé

Abstract

Finding optimal measurement schemes in quantum state tomography is a fundamental problem in quantum computation. It is known that for non-degenerate operators the optimal measurement scheme is based on mutually unbiassed bases. This paper is a follow up from our previous work, where we use standard numerical approaches to look for optimal measurement schemes, where the measurement operators are projections on individual pure quantum states. In this paper we demonstrate the usefulness of several machine learning techniques – reinforcement learning and parallel machine learning approaches, to discover measurement schemes, which are significantly better than the ones discovered by standard numerical methods in our previous work. The high-performing quorums of projection operators we have discovered have complex structure and symmetries, which may imply that the optimal solution will possess such symmetries.

Mots clés

  • Quantum information
  • optimization problem
  • Widening
  • reinforcement learning
Accès libre

A New Class of “Growth Functions” with Polynomial Variable Transfer Generated by Real Reaction Networks

Publié en ligne: 31 Dec 2020
Pages: 74 - 81

Résumé

Abstract

In [4, 5], two classes of growth models with “exponentially variable transfer” and “correcting amendments of Bateman-Gompertz-Makeham-type” based on a specific extended reaction network have been studied [1]. In this article we will look at the new scheme with “polynomial variable transfer”. The consideration of such a dynamic model in the present article is dictated by our passionate desire to offer an adequate model with which to well approximate specific data in the field of computer viruses propagation, characterized by rapid growth in the initial time interval. Some numerical examples, using CAS Mathematica illustrating our results are given.

Mots clés

  • Reaction networks
  • Generalized growth model
  • Exponentially and polynomial variable transfers
Accès libre

Comments on the Half Logistic Inverse Rayleigh (HLIR) cdf with “Polynomial Variable Transfer”. Some Applications

Publié en ligne: 31 Dec 2020
Pages: 82 - 93

Résumé

Abstract

On the base of the Half Logistic – G family of distributions proposed by Cordeiro, Alizadeh and Marinho [2] some mathematical properties are investigated by Almarashi et al. [1]. We study the “saturation” to the horizontal asymptote: t=1 by the new growth function M(t) in the Hausdorff sense. Similar to our previous studies [3-6], in this article we will define and analyze in detail the new family. We will call this family the “Half-Logistic-Inverse-Rayleigh cdf with Polynomial Variable Transfer” (HLIRPVT) cdf. Section 3 shows the potentiality of proposed new model under four real data sets. Some numerical examples using CAS MATHEMATICA are given.

Mots clés

  • Half-Logistic-Inverse-Rayleigh (HLIR) cdf
  • Half-Logistic-Inverse-Rayleigh cdf with Polynomial Variable Transfer (HLIRPVT) cdf
  • “Saturation” in Hausdorff sense
  • Hausdorff distance
  • Upper and lower bounds
Accès libre

Performance Analysis of a Scalable Algorithm for 3D Linear Transforms on Supercomputer with Intel Processors/Co-Processors

Publié en ligne: 31 Dec 2020
Pages: 94 - 104

Résumé

Abstract

Practical realizations of 3D forward/inverse separable discrete transforms, such as Fourier transform, cosine/sine transform, etc. are frequently the principal limiters that prevent many practical applications from scaling to a large number of processors. Existing approaches, which are based primarily on 1D or 2D data decompositions, prevent the 3D transforms from effectively scaling to the maximum (possible/available) number of computer nodes. A highly scalable approach to realize forward/inverse 3D transforms has been proposed. It is based on a 3D decomposition of data and geared towards a torus network of computer nodes. The proposed algorithms requires compute-and-roll time-steps, where each step consists of an execution of multiple GEMM operations and concurrent movement of cubical data blocks between nearest neighbors. The aim of this paper is to present an experimental performance study of an implementation on high performance computer architecture.

Mots clés

  • 3D linear transforms
  • parallel implementation
  • Intel processors/ co-processors
Accès libre

Numerical Solution of Integro-Differential Equations Modelling the Dynamic Behavior of a Nano-Cracked Viscoelastic Half-Plane

Publié en ligne: 31 Dec 2020
Pages: 105 - 115

Résumé

Abstract

The scattering of time-harmonic waves by a finite, blunt nano-crack in a graded, viscoelastic bulk material with a free surface is considered in this work. Non-classical boundary conditions and a localized constitutive equation at the interface between crack and matrix, following the Gurtin-Murdoch surface elasticity theory are introduced. An efficient numerical technique is developed using integro-differential equations along the nano-crack line that is based on an analytically derived Green‘s function for the quadratically inhomogeneous half-plane. The dependence of the diffracted and scattered waves and of the local stress concentration fields on key problem parameters such as viscosity, inhomogeneity, surface elasticity, and interaction between the nano-crack and the free surface are all examined through an extensive parametric study.

Mots clés

  • Viscoelasticity
  • graded half-plane
  • nano-crack
  • integro-differential equations
  • stress concentration
  • wave scattering
  • wave diffraction
Accès libre

On Dynamic Parallelization of Multilevel Monte Carlo Algorithm

Publié en ligne: 31 Dec 2020
Pages: 116 - 125

Résumé

Abstract

MultiLevel Monte Carlo (MLMC) attracts great interest for numerical simulations of Stochastic Partial Differential Equations (SPDEs), due to its superiority over the standard Monte Carlo (MC) approach. MLMC combines in a proper manner many cheap fast simulations with few slow and expensive ones, the variance is reduced, and a significant speed up is achieved. Simulations with MC/MLMC consist of three main components: generating random fields, solving deterministic problem and reduction of the variance. Each part is subject to a different degree of parallelism. Compared to the classical MC, MLMC introduces “levels” on which the sampling is done. These levels have different computational cost, thus, efficiently utilizing the parallel resources becomes a non-trivial problem. The main focus of this paper is the parallelization of the MLMC Algorithm.

Mots clés

  • SPDE
  • MLMC
  • UQ
  • Parallelization
  • Flow in random porous media
Accès libre

Geometrically Non-Linear Vibration of a Cantilever Interacting with Rarefied Gas Flow

Publié en ligne: 31 Dec 2020
Pages: 126 - 139

Résumé

Abstract

The work is devoted to study 2D pressure driven rarefied gas flow in a microchannel having an elastic obstacle. The elastic obstacle is clamped at the bottom channel wall and its length is half of the channel height. The gas flow is simulated by Direct Simulation Monte Carlo (DSMC) method applying the advanced Simplified Bernoulli Trial (SBT) collision scheme. The elastic obstacle is modelled as geometrically nonlinear Euler Bernoulli beam. A reduced 3 modes reduction model of the beam is created. The influence of the gas flow on the beam vibration is studied, considering the linear and nonlinear beam theories.

Mots clés

  • Fluid-Structure Interaction (FSI)
  • rarefied gas
  • microchannel
  • DSMC
  • Euler-Bernoulli beam
  • pressure driven flow
Accès libre

Molecular Dynamics Simulations of Acetylcholinesterase – Beta-Amyloid Peptide Complex

Publié en ligne: 31 Dec 2020
Pages: 140 - 154

Résumé

Abstract

Alzheimer’s Disease (AD) is a neurodegenerative disorder with severe consequences and lethal outcome. One of the pathological hallmarks of the disease is the formation of insoluble intercellular beta-Amyloid (Aβ) plaques. The enzyme ACetylcholinEsterase (AChE) promotes and accelerates the aggregation of toxic Aβ protofibrils progressively converted into plaques. The Peripheral Anionic Site (PAS), part of the binding gorge of AChE, is one of the nucleation centers implicated in the Aβ aggregation. In this study, the Aβ peptide was docked into the PAS and the stability of the formed complex was investigated by molecular dynamics simulation for 1 μs (1000 ns). The complex was stable during the simulation. Apart from PAS, the Aβ peptide makes several additional contacts with AChE. The main residence area of Aβ on the surface of AChE is the region 344-361. This region is next to PAS but far enough to be sterically hindered by dual-site binding AChE inhibitors.

Mots clés

  • Alzheimer’s disease
  • beta-Amyloid (Aβ)
  • acetylcholinesterase (AChE)
  • Peripheral Anionic Site (PAS)
  • Aβ aggregation
  • senile plaques
  • amyloid fibrils
  • neurodegenerative disorder
  • molecular dynamics
Accès libre

Trend Analysis of CMIP5 Ensemble of Climate Indices over Southeast Europe with Focus on Agricultural Impacts

Publié en ligne: 31 Dec 2020
Pages: 155 - 165

Résumé

Abstract

Nowadays there is a strong degree of agreement that the climate change is the defining challenge of our time, which will exert influence on the ecosystems, on all branches of the international economy and on the quality of life. The analysis based on climate indices is widely used non-parametric approach for quantification of the mean state as well as extreme climate events. This study, which is continuation of our previous efforts, is dedicated to the assessment of the trend magnitude and the trend statistical significance of six temperature-based and three precipitation-based indices in projected future climate over Southeast Europe up to the end of the 21st century. The indices are computed from the bias-corrected output of five CMIP5 global models, reinforced with all four RCP emission scenarios. The model output is accessed from the section of the Inter Sectoral Impact Model Intercomparison Project in the Copernicus Data Store. The multi model ensemble medians of the temperature-based indices shows considerable increase which is consistent with the warming of the mean temperatures. These changes are statistically significant in most cases and intensify with the radiative forcing. The revealed tendencies of the precipitation-based indices are more complex when compared with temperature tendencies.

Mots clés

  • Climate indices
  • RCP
  • CMIP5 Ensemble
  • future climate
  • trend analysis
Accès libre

Degree-Day Climatology over Central and Southeast Europe for the Period 1961-2018 – Evaluation in High Resolution

Publié en ligne: 31 Dec 2020
Pages: 166 - 174

Résumé

Abstract

The ongoing climate change over Central and Southeast Europe has a great potential to affect significantly the public energy demands and in particular the energy consumption in the residential heating and cooling sector. The linkage of the ambient daily extreme and mean temperatures and the energy needs for condition or heat buildings can be quantified as numerical indicators as the heating and cooling degree-days. In the present study, these indicators are calculated according the UK Met Office methodology from the daily mean and extreme temperatures, which, in turn, are computed from the output of the MESCAN-SURFEX system in the frame the FP7 UERRA project. The study, which is performed in a very high resolution, is dedicated on the analysis of the spatial patterns as well as assessment of the magnitude and statistical significance of the temporal evolution of the heating and cooling degree-days. It reveals general tendencies which are coherent with the regional climate warming, but with high spatial heterogeneities. The study confirms the essential impact of the ongoing climate change on the heating, ventilating and air-conditioning industry over Central and Southeast Europe.

Mots clés

  • Degree-days
  • MESCAN-SURFEX
  • UERRA Project
  • trend analysis
Accès libre

AllerScreener – A Server for Allergenicity and Cross-Reactivity Prediction

Publié en ligne: 31 Dec 2020
Pages: 175 - 184

Résumé

Abstract

Allergenicity of proteins is a subtle property encoded in their structures. The prediction of allergenicity of novel proteins saves time and resources for subsequent experimental work. In the host antigen-presenting cells, the allergens are processed as antigens by the means of Human Leukocyte Antigens (HLA) class II proteins. Sometimes, people allergic to a given protein show allergic reaction to a different protein, even when the two proteins have different routes of exposure. This phenomenon is termed cross-reactivity. Here, we describe a server for allergenicity and cross-reactivity prediction based on the abilities of allergenic proteins to generate binders to HLA class II proteins. The generated peptides are compared to HLA binders originating from known allergens. As a result, the server returns a list of common binders, origin proteins, and species. Different species generate common HLA binders and this determines their cross-reactivity. The server is named AllerScreener and is freely accessible at: http://www.ddg-pharmfac.net/AllerScreener.

Mots clés

  • AllerScreener
  • allergenicity
  • cross-reactivity
  • EpiTOP
  • EpiDOCK
  • HLA class II
Accès libre

Design of Multi-Epitope Vaccine against SARS-CoV-2

Publié en ligne: 31 Dec 2020
Pages: 185 - 193

Résumé

Abstract

The ongoing COVID-19 pandemic requires urgently specific therapeutics and approved vaccines. Here, the four structural proteins of the Severe Acute Respiratory Syndrome CoronaVirus 2 (SARS-CoV-2), the causative agent of COVID-19, are screened by in-house immunoinformatic tools to identify peptides acting as potential T-cell epitopes. In order to act as an epitope, the peptide should be processed in the host cell and presented on the cell surface in a complex with the Human Leukocyte Antigen (HLA). The aim of the study is to predict the binding affinities of all peptides originating from the structural proteins of SARS-CoV-2 to 30 most frequent in the human population HLA proteins of class I and class II and to select the high binders (IC50 < 50 nM). The predicted high binders are compared to known high binders from SARS-CoV conserved in CoV-2 and 77% of them coincided. The high binders will be uploaded onto lipid nanoparticles and the multi-epitope vaccine prototype will be tested for ability to provoke T-cell mediated immunity and protection against SARS-CoV-2.

Mots clés

  • COVID-19
  • multi-epitope vaccine
  • SARS-CoV
  • SARS-CoV-2
  • EpiJen
  • EpiTOP
  • EpiDOCK
  • high binders
  • HLA class I
  • HLA class II
Accès libre

Immunoinformatic Analysis of Human Thyroglobulin

Publié en ligne: 31 Dec 2020
Pages: 194 - 200

Résumé

Abstract

The AutoImmune ThyroiDitis (AITD), known as Hashimoto’s disease, is a chronic autoimmune thyroid disease progressively developed to hypothyroidism. The AITD is characterized by the formation of autoantibodies targeting two specific thyroid antigens, Thyroglobulin (Tg) and Thyroid PerOxidase (TPO). Tg is a precursor of the thyroid hormones while TPO catalyses their synthesis. The AITD has a strong genetic predisposition. During the last years, it was found that the susceptibility to AITD is associated with certain Human Leukocyte Antigens (HLA) class II genes of loci DR and DQ. In the present study, we applied in-house immunoinformatic tools to identify peptides originating from Tg and binding to AITD susceptible alleles: HLA-DR3, HLA-DR4, HLA-DR5, HLA-DQ2 and HLA-DQ8. Five peptide fragments containing promiscuous overlapping binders were selected. These were p470, p949, p1948, p2348 and p2583. Only one of them contains a known epitope (p1948). The rest have not been reported yet. The selected peptide fragments will be coupled to monoclonal antibodies specific to inhibitory B cell receptors designed to suppress the production of Tg autoantibodies.

Accès libre

Finite-Temperature Single Molecule Vibrational Dynamics from Combined Density Functional Tight Binding Extended Lagrangian Dynamics Simulations and Time Series Analysis

Publié en ligne: 31 Dec 2020
Pages: 201 - 212

Résumé

Abstract

Combining a computationally efficient and affordable molecular dynamics approach, based on atom-centered density matrix propagation scheme, with the density functional tight binding semiempirical quantum mechanics, we study the vibrational dynamics of a single molecule at series of finite temperatures, spanning quite wide range. Data generated by molecular dynamics simulations are further analyzed and processed using time series analytic methods, based on correlation functions formalism, leading to both vibrational density of states spectra and infrared absorption spectra at finite temperatures. The temperature-induced dynamics in structural intramolecular parameters is correlated to the observed changes in the spectral regions relevant to molecular detection. In particular, we consider a case when an intramolecular X-H stretching vibrational states are notably dependent on the intramolecular torsional degree of freedom, the dynamics of which is, on the other hand, strongly temperature-dependent.

Mots clés

  • Atom-centered density matrix propagation
  • density functional tight binding
  • molecular dynamics
  • finite-temperature vibrational dynamics
  • formic acid
  • torsional motion

Planifiez votre conférence à distance avec Sciendo