A scientific on-line database for efficient function approximation

Authors:
Ivana Veljkovic;Paul E. Plassmann;Daniel C. Haworth
Affiliations:
Department of Computer Science and Engineering, The Pennsylvania State University, University Park, PA;Department of Computer Science and Engineering, The Pennsylvania State University, University Park, PA;Department of Mechanical and Nuclear Engineering, The Pennsylvania State University, University Park, PA
Venue:
ICCSA'03 Proceedings of the 2003 international conference on Computational science and its applications: PartI
Year:
2003

Citing 0
Cited 4

Indexing for function approximation

VLDB '06 Proceedings of the 32nd international conference on Very large data bases
Scalable photon monte carlo algorithms and software for the solution of radiative heat transfer problems

HPCC'05 Proceedings of the First international conference on High Performance Computing and Communications
A scalable scientific database for chemistry calculations in reacting flow simulations

HPCC'05 Proceedings of the First international conference on High Performance Computing and Communications
Parallel heuristics for an on-line scientific database for efficient function approximation

PARA'04 Proceedings of the 7th international conference on Applied Parallel Computing: state of the Art in Scientific Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

The use of a scientific, on-line database for function approximation is a technique that can significantly reduce the computational demands for problems requiring the frequent evaluation of computationally expensive functions. In this paper we present several new algorithms which significantly improve the performance of software implementations of this database approach. These algorithms are of two types--first, algorithms designed to improve the database retrieval rates; and, second, algorithms that seek to reduce the size of the database and subsequently reduce the cost of database queries. We have developed a software implementation of algorithms called DOLFA. We present experimental results which detail the performance of the DOLFA software for one representative combustion application. We observe significant improvements in cumulative time needed for database operations and memory requirements in a comparison of DOLFA to the function tabulation software system ISAT.