Abstract
We consider a method for constrained minimization based on recursive quadratic programming In this note we suggest an alternative form to obtain a descent direction for the penalty function. Some computational results are also presented.
C.R Categories: