Abstract
A FORTRAN subroutine for locating the minimum of a function is presented. The subroutine uses a simplex design and iterative searching strategy similar to the Nelder-Mead strategy. When the path to the optimum is reasonably straight, the proposed strategy converges with fewer iterations than the Nelder-Mead strategy, particularly with more than four parameters.
Additional information
Notes on contributors
David K. Smith
Mr. Smith is a Member of the Mathematical and Statistical Analysis Group in the Research and Development Division.