Answer :
Solver in nonlinear programming can indeed stop at a local optima instead of the global optimum due to the presence of multiple local optima and the limitations of the searching algorithm. So, the correct answer is option a) True.
The statement, "For NLP (Nonlinear Programming) problems, Solver may stop at a point that is a local rather than a global optimum," is true. This is because a response surface can have multiple local optima, and only one of these will be the global optimum. When using Solver, the algorithm may stop the search when the objective function's value improves only marginally. This stopping point is determined by the convergence criterion, which users can adjust. However, there is no guarantee that the stopping point is the global optimum; it could very well be a local optimum instead.
If a search starts near a local optimum, the algorithm might not reach the global optimum. This is representatively shown with a ridge having several peaks, and only a search starting at the correct position will reach the global optimum. A poorly designed algorithm, or one with limitations in the possible directions it can explore, might encounter difficulties reaching the global optimum, instead settling for a local one. So, the correct answer is option a) True.
