Conditioning of Quasi-Newton Methods for Function Minimization
Reads0
Chats0
TLDR
In this paper, a class of approximating matrices as a function of a scalar parameter is presented, where the problem of optimal conditioning of these matrices under an appropriate norm is investigated and a set of computational results verifies the superiority of the new methods arising from conditioning considerations to known methods.Abstract:
Quasi-Newton methods accelerate the steepest-descent technique for function minimization by using computational history to generate a sequence of approximations to the inverse of the Hessian matrix. This paper presents a class of approximating matrices as a function of a scalar parameter. The problem of optimal conditioning of these matrices under an appropriate norm as a function of the scalar parameter is investigated. A set of computational results verifies the superiority of the new methods arising from conditioning considerations to known methods.read more
Citations
More filters
Proceedings ArticleDOI
An Improved Analysis of (Variance-Reduced) Policy Gradient and Natural Policy Gradient Methods
TL;DR: This paper revisits and improves the convergence of policy gradient, natural PG (NPG) methods, and their variance-reduced variants, under general smooth policy parametrizations, and proposes SRVR-NPG, which incorporates variancereduction into the NPG update.
Journal ArticleDOI
Theoretical study of the proton affinities of 2-, 3-, and 4-monosubstituted pyridines in the gas phase by means of MINDO/3, MNDO, and AM1: Proton Affinities of Pyridines
TL;DR: In this paper, the MINDO/3, MNDO, and AM1 methods were compared with experimental results, and it was found that all the MMDO methods are ca. 6% too high (mean value) compared to the experimental results; however, a much better agreement has been observed for the AM1 method where the theoretical values are only ca. 2.4% too low (measured mean value).
Journal ArticleDOI
A survey of optimization procedures for stable structures and transition states
TL;DR: In this paper, the authors examined a variety of methods for obtaining the stable geometry of molecules and the transition states of simple systems and summarized some of their findings. They found the most efficient methods for optimizing structure to be those based on calculated gradients and estimated second derivative (Hessian) matrices, the later obtained either from the Broyden-Fletcher-Goldfarb-Shanno (BFGS) quasi-Newton update method or from approximations to the coupled perturbed Hartree-Fock method.
Journal ArticleDOI
Electrochemical and spectroscopic investigation of the influence of acid-base chemistry on the redox properties of 2,5-dimercapto-1,3,4-thiadiazole
TL;DR: In this article, the effect of acid-base processes on the redox behavior of 2,5-dimercapto-1,3,4-thiadiazole (DMcT) was studied in the absence and presence of pyridine (Py) or triethylamine (TEA) in acetonitrile solutions.
Book ChapterDOI
Numerical optimization techniques
TL;DR: The first formal statement of nonlinear programming (numerical optimization) applied to structural design was offered by Schmit in 1960 and has evolved at an ever-increasing pace until it can now be considered to be reasonably mature.
References
More filters
Journal ArticleDOI
A Rapidly Convergent Descent Method for Minimization
Roger Fletcher,M. J. D. Powell +1 more
TL;DR: A number of theorems are proved to show that it always converges and that it converges rapidly, and this method has been used to solve a system of one hundred non-linear simultaneous equations.
Journal ArticleDOI
A family of variable-metric methods derived by variational means
TL;DR: In this paper, a rank-two variable-metric method was derived using Greenstadt's variational approach, which preserves the positive-definiteness of the approximating matrix.
Journal ArticleDOI
A Class of Methods for Solving Nonlinear Simultaneous Equations
TL;DR: In this article, the authors discuss certain modifications to Newton's method designed to reduce the number of function evaluations required during the iterative solution process of an iterative problem solving problem, such that the most efficient process will be that which requires the smallest number of functions evaluations.
Journal ArticleDOI
Quasi-Newton methods and their application to function minimisation
TL;DR: The Newton-Raphson method as mentioned in this paper is one of the most commonly used methods for solving nonlinear problems, where the corrections are computed as linear combinations of the residuals.
Journal ArticleDOI
A Comparison of Several Current Optimization Methods, and the use of Transformations in Constrained Problems
TL;DR: Transitions whereby inequality constraints of certain forms can be eliminated from the formulation of an optimization problem are described, and examples of their use compared with other methods for handling such constraints are described.