On line search termination criteria for and convergence of conic algorithms for minimization
Conference
·
OSTI ID:5606031
This report presents an implementable class of collinear scaling algorithms for unconstrained minimization. A member of this class that extends the BFGS method does possess desirable properties similar to those shown for the BFGS method by Powell (1976). However, a major drawback is that in order to guarantee q-superlinear convergence, a parameter in the algorithm must be set to a value which depends on F. 8 refs.
- Research Organization:
- Washington State Univ., Pullman (USA). Dept. of Pure and Applied Mathematics
- DOE Contract Number:
- FG06-85ER25007
- OSTI ID:
- 5606031
- Report Number(s):
- CONF-8610302-1; ON: DE88002844
- Country of Publication:
- United States
- Language:
- English
Similar Records
Algorithms with conic termination for nonlinear optimization
A tool for the analysis of quasi-Newton methods with application to unconstrained minimization
The global convergence of partitioned BFGS on problems with convex decompositions and Lipschitzian gradients
Technical Report
·
Mon Nov 30 23:00:00 EST 1987
·
OSTI ID:5667958
A tool for the analysis of quasi-Newton methods with application to unconstrained minimization
Technical Report
·
Mon Nov 30 23:00:00 EST 1987
·
OSTI ID:5470847
The global convergence of partitioned BFGS on problems with convex decompositions and Lipschitzian gradients
Technical Report
·
Mon Nov 30 23:00:00 EST 1987
·
OSTI ID:5434080