Computational experience with improved variable metric methods for unconstrained minimization

Institute of Mathematics of the Academy of Sciences of the Czech Republic provides access to digitized documents strictly for personal use. Each copy of any part of this document must contain these Terms of use. This paper has been digitized, optimized for electronic delivery and stamped with digital signature within the project DML-CZ: The Czech Digital Mathematics Library The paper describes three improved variable metric methods for unconstrained minimization and shows their efficiency on a broad class of test problems. These methods are based on the controlled scaling and on the pertinent combination of the rank-one method with other variable metric methods.