It is demonstrated that substantial savings in the computer storage space and calculation can be effected by a close inspection of the subsearch phase of a multivariate optimization algorithm. This is particularly important in implementing algorithms on small computers. A significant compaction of the golden section search is developed. Coincidentally, it is found that the actual bracketing and convergence properties are better than those conventionally used in the literature.

This content is only available via PDF.
You do not currently have access to this content.