Computer Science Journal (AGH University of Science and Technology, Krakow)
Not a member yet
476 research outputs found
Sort by
MODELING AND RENDERING OF CONVECTIVE CUMULUS CLOUDS FOR REAL-TIME GRAPHICS PURPOSES
Get PDFThe paper presents a simulation and rendering model of three dimensional covective cloud evolution. The model is physically based, however its purpose is graphical. The main stress is put on balancing two parts of a model: the atmsphere simulation with convective motion of air and water vapor combined with rendering of semi-transparent and light-scattering clouds, in order to achieve realistic animation in real-time. We examine and compare two algorithmic approaches based on CPU and GPU computations
Tree Structures for Adaptive Control Space in 3D Meshing
Get PDFThe article presents a comparison of several octree- and kd-tree-based structuresused for the construction of control space in the process of anisotropicmesh generation and adaptation. The adaptive control space utilized by theauthors supervises the construction of meshes by providing the required metricinformation regarding the desired shape and size of elements of the mesh ateach point of the modeled domain. Comparative tests of these auxiliary structureswere carried out based on different versions of the tree structures withrespect to computational and memory complexity as well as the quality of thegenerated mesh. Analysis of the results shows that kd-trees (not present inthe meshing literature in this role) offer good performance and may becomea reasonable alternative to octree structures
The Influence of Listener Personality on Music Choices
Get PDFTo deliver better recommendations, music information systems need to go beyondstandard methods for musical taste prediction. Tracking the listener’semotions is one way to improve the quality of recommendations. This can beachieved explicitly, by asking the listener to report his/her emotional state, orimplicitly, by tracking the context in which the music is heard. However, thefactors that induce particular emotions vary among individuals. This paperpresents initial research on the influence of an individual’s personality on hisor her choice of music. The psychological profile of a group of 16 students wasdetermined by a questionnaire. The participants were asked to label their ownmusic collections, listen to the music, and mark their emotions using a customapplication. Statistical analysis revealed correlations between low-level audiofeatures, personality types, and emotional states of the students
Applications of Alternating Direction Solver for simulations of time-dependent problems
Get PDFThis paper deals with application of Alternating Direction solver (ADS) to nonstationarylinear elasticity problem solved with isogeometric FEM. Employingtensor product B-spline basis in isogeometric analysis under some restrictionsleads to system of linear equations with matrix possessing tensor product structure.Alternating Direction Implicit algorithm is a direct method that exploitsthis structure to solve the system in O (N ), where N is a number of degreesof freedom (basis functions). This is asymptotically faster than state-of-theartgeneral purpose multi-frontal direct solvers. In this paper we also presentthe complexity analysis of ADS incorporating dependence on order of B-splinebasis
Using Redis supported by NVRAM in HPC applications
Get PDFNowadays, the efficiency of a storage systems is a bottleneck in many moern HPC clusters. High performance in traditional approach – processing using files – is often difficult to obtain because of model complexity and its read/write patterns. Alternative approach is applying a key-value database, which usually has low latency and scales well. On the other hand, many key-value stores suffer from limitation of memory capacity and vulnerability to serious faiures, which is caused by processing in RAM. Moreover, some research suggests, that scientific data models are not applicable to storage structures of key-value databases. In this paper, the author proposes resolving mentioned issues by replacing RAM with NVRAM. Practical example is based on Redis NoSQL. The article contains also a three domain specific APIs, that show the idea bhind transformation from HPC data model to Redis structures, as well as two micro-benchmarks results
The Numerical Approaches to Heat Transfer Problem in Modern Electronic Structures
No full textThe main aim of this paper is to present a detailed description of the research related to the modeling of heat conduction in modern electronic structures including special consideration of numerical aspects of analyzed algorithms.The motivation to undertake the research and some most important results of experiments and simulations are also included.Moreover, the numerical appoximation of the problem as well as the methodology and the sample solution of mentioned problem are presented.In the main part, the discretisation techniques, Ordinary Differential Equation algorithms and simulation results for Runge-Kutta and Gear\u27s algorithms are analysed and discussed.Additionally, a new effective approach to modelling of heat transfer in electronic nanostructures is demonstrated
Building Synchrotron ontology: the analysis of synchrotron control system in collaborative environment
Get PDFThis paper presents research on building a synchrotron ontology using analysis of a synchrotron control system and the aspects of collaborative ontology engineering. It includes a general description of the studied domain and the method used to develop a synchrotron domain ontology. The ontology is being created on the basis of a Solaris synchrotron control system in cooperation with the synchrotron facilities belonging to the Tango community. The first Polish synchrotron radiation facility Solaris is located at Jagiellonian University’s Third Campus in Krakow, Poland. Synchrotron is an unique source of electromagnetic radiation known as synchrotron radiation. This paper discusses the impact of the Solaris control system on the building of a synchrotron ontology. It also includes the main assumptions relating to the collaborative knowledge acquired for this domain. The synchrotron ontology will support the optimization of existing control systems and the development of a new synchrotron control system based on Tango controls or other technologies in a consistent manner. Using the same general assumptions and terms, this could be later used for integration and data sharing purposes. The synchrotron ontology can facilitate interoperation by the integration of information from different sources from one or many synchrotron control systems and integrate different parts of the controls systems that provide analogical or similar services. It can also be used to support the translation between different representations, especially regarding particular devices. Knowledge sharing and reuse is a big challenge in complex, distributed systems where the knowledge required is very specialized for different sets of functionalities or subsystems. Regarding synchrotron systems, many specialists must provide their support so the IT specialists are able to develop and maintain a control system. In this case, the synchrotron ontology can be a guideline for knowledge sharing and reuse
Analysis of annual temperature distribution inside the experimental embankment
Get PDFThe aim of this paper is analysis of temperatures distribution inside experimental embankment from August 2015 to September 2016. Analysis was carried out in order to interpretation of the results of the flood experiments performed on the experimental embankment. The reference for year temperature changes in the embankment at various depths was obtained. A simplified model of temperature changes depending on the depth was made. The model parameters which can be used for modelling the temperatures in the embankment during the experiments were estimated
APPLICATION OF MULTI-CRITERIA ANALYSIS BASED ON THE INDIVIDUAL PSYCHOLOGICAL PROFILE FOR RECOMMENDER SYSTEMS
Get PDFThis paper presents a novel approach for user classification exploiting multicriteriaanalysis. This method is based on measuring the distance between anobservation and its respective Pareto front. The obtained results show that thecombination of the standard KNN classification and the distance from Paretofronts gives satisfactory classification accuracy – higher than the accuracy obtainedfor each of these methods applied separately. Conclusions from thisstudy may be applied in recommender systems where the proposed methodcan be implemented as the part of the collaborative filtering algorithm
Hardware aware tiling optimization for multi-core systems
Get PDFThis paper presents a proposition of the new tool which improves tiling efficiencyfor given hardware architecture. This article also describes the correlationbetween changing hardware architecture and methods of software optimization.First chapter includes short description of the change in hardware architecturewhich has occurred in recent 10 years. The second chapter provides an overviewof tools which will be used in further research. The consecutive sections containdescription of proposed hardware-aware tool for optimal tiling