Efficient computation of spectral bounds for Hessian matrices on hyperrectangles for global optimization

Abstract: We compare two established and a new method for the calculation of spectral bounds for Hessian matrices on hyperrectangles by applying them to a large collection of 1522 objective and constraint functions extracted from benchmark global optimization problems. Both the tightness of the spectral bounds and the computational effort are assessed. Specifically, we compare eigenvalue bounds obtained with the interval variant of Gershgorin's circle criterion [2,6], Hertz and Rohn's [7,16] method for tight bounds of interval matrices, and a recently proposed Hessian matrix eigenvalue arithmetic [12], which deliberately avoids the computation of interval Hessians.

Review text

Let \(\varphi :U\rightarrow {\mathbb R}\) be a twice continuously differentiable function on an open set \(U\subseteq {\mathbb R}^n\) and let \( B = [\underline x_1, \overline x_1]\times \ldots \times [\underline x _i, \overline x_n ] \) be a closed hyperrectangle in \(U\). The present paper concerns the following problem \[ \begin{aligned}& \text{Find }\underline \lambda\in {\mathbb R}, \overline \lambda \in {\mathbb R} \text{ such that }\\ &\underline \lambda\leq \lambda \leq \overline \lambda \text{ for all eigenvalues } \lambda \text{ of all matrices } H \in H (\varphi, B), \end{aligned} \] where \( H (\varphi, B)\) is the set of Hessian matrices of \(\varphi\) on B, \[ H (\varphi, B) = \{ \nabla^2 \varphi (x): x \in B \}. \] A bound \(\overline \lambda\) (resp. \(\underline \lambda\)) is called tight if there exists at least one matrix \(H\) in the matrix set with an eigenvalue \(\lambda = \overline \lambda\) (resp. \(\lambda= \underline \lambda\)). Note that the bounds \(\underline \lambda, \overline \lambda\) above may or may not be tight. The authors compare two established and a new methods for the calculation of spectral bounds for Hessian matrices on hyperrectangles by applying them to a large collection of 1,522 objective and constraint functions extracted from benchmark global optimization problems. Both the tightness of the spectral bounds and the computational effort of the three methods, which apply to \(C^2\) functions \(\varphi : {\mathbb R}\rightarrow {\mathbb R}\) that can be written as codelists, are assessed.

Recommendations

Cites work

Cited in

(6)

Describes a project that uses

Uses Software

COCONUT

This page was built for publication: Efficient computation of spectral bounds for Hessian matrices on hyperrectangles for global optimization

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2250109)