Accession Number:

ADA207718

Title:

Large Scale Function Minimization

Descriptive Note:

Final scientific rept. 15 Jul 1985-14 Oct 1988

Corporate Author:

JOHNS HOPKINS UNIV BALTIMORE MD DEPT OF MATHEMATICAL SCIENCES

Personal Author(s):

Report Date:

1988-10-14

Pagination or Media Count:

5.0

Abstract:

Several new optimization techniques suitable for parallel computers have been developed and have been tested on an Intel hypercube. On a number of nonlinear problems, the algorithms tested have demonstrated dramatic speed-ups over their sequential counterparts. The fact that these speedups are better than can be attributed to parallelism done suggest that they may lead to improved sequential methods.

Subject Categories:

  • Theoretical Mathematics

Distribution Statement:

APPROVED FOR PUBLIC RELEASE