Derivative-free optimization for parameter estimation in computational nuclear physics

TL;DR

This paper explores derivative-free optimization methods, specifically POUNDERS, for estimating parameters in nuclear density functional theory, demonstrating its effectiveness on complex calibration problems without relying on derivatives.

Contribution

It introduces the application of the POUNDERS derivative-free optimizer to nuclear physics parameter estimation, highlighting its efficiency and consistency in complex, derivative-free calibration tasks.

Findings

POUNDERS achieves consistent solutions across various calibration problems.

Derivative-free optimization is effective for computationally expensive nuclear models.

The paper provides a practical guide to using POUNDERS in nuclear physics applications.

Abstract

We consider optimization problems that arise when estimating a set of unknown parameters from experimental data, particularly in the context of nuclear density functional theory. We examine the cost of not having derivatives of these functionals with respect to the parameters. We show that the POUNDERS code for local derivative-free optimization obtains consistent solutions on a variety of computationally expensive energy density functional calibration problems. We also provide a primer on the operation of the POUNDERS software in the Toolkit for Advanced Optimization.