The recursive construction of Schur-complements is used to construct a mult
i-level preconditioner for an iterative linear solver. For each level, the
removed unknowns are selected in such a way that the eliminated matrix bloc
k is strictly diagonally dominant. A Newton-type iteration scheme is used t
o construct a sparse approximate inverse of this sub-matrix. The threshold
for the diagonal dominance controls the computational effort to achieve a c
ertain accuracy in the Newton iteration. We present a modification of the g
reedy algorithm in order to identify a suitable sub-matrix that is diagonal
ly dominant and ensures a stable forward and backward substitution. Some ex
amples are presented. Copyright (C) 2000 John Wiley & Sons, Ltd.