A subspace, interior, and conjugate gradient method for large-scale bound-constrained minimization problems

A subspace adaptation of the Coleman-Li trust region and interior method is proposed for solving large-scale bound-constrained minimization problems. This method can be implemented with either sparse Cholesky factorization or conjugate gradient computation. Under reasonable conditions the convergenc e properties of this subspace trust region method are as strong as those of its full-space version. Computational performance on various large test problems is reported; advan tages of our approach are demonstrated. Our experience indicates that our p roposed method represents an efficient way to solve large bound-constrained minimization problems.