Skip to search formSkip to main contentSkip to account menu

Symmetric rank-one

Known as: SR1 formula 
The Symmetric Rank 1 (SR1) method is a quasi-Newton method to update the second derivative (Hessian)based on the derivatives (gradients) calculated… 
Wikipedia (opens in a new tab)

Papers overview

Semantic Scholar uses AI to extract papers important to this topic.
2014
2014
The problem of symmetric rank‐one approximation of symmetric tensors is important in independent components analysis, also known… 
2011
2011
In this paper, we investigate a symmetric rank-one (SR1) quasi-Newton (QN) formula in which the Hessian of the objective function… 
2011
2011
We propose a quasi-Newton line-search method that uses negative curvature directions for solving unconstrained optimization… 
2010
2010
In this paper we investigate on convergence rate of a modified symmetric rank-one (SR1) method for unconstrainedoptimization… 
Review
2010
Review
2010
The focus of this thesis is on analyzing the theoretical and computational aspects of some quasi-Newton (QN) methods for locating… 
2009
2009
In this paper, we present a generalized Symmetric Rank-one (SR1) method by employing interpolatory polynomials in order to… 
2009
2009
Problem statement: Memoryless QN methods have been regarded effective techniques for solving large-scale problems that can be… 
1998
1998
The use of the symmetric rank-one Hessian update and the Ž . Broyden]Fletcher]Goldfarb]Shano BFGS update formula are considered… 
1997
1997
Metric-based SR1 updates which are stabilized by a variationalrelaxation of the quasi-Newton relation are examined… 
1995
1995
We consider conditions under which the SR1 iteration is locally convergent. We apply the result to a pointwise structured SR1…