Second-Order Cone Programming for P-Spline Simulation Metamodeling

TL;DR

This paper introduces a novel method for approximating simulation models using nonnegative B-splines with a penalty on high-order differences, formulated as a second-order cone programming problem for efficient solution.

Contribution

It presents a new approach that combines P-spline approximation with second-order cone programming to ensure nonnegativity and prevent overfitting in simulation metamodeling.

Findings

Efficient solution via modern optimization techniques.

Implementation available in MATLAB/GNU Octave.

Improved approximation accuracy with nonnegative constraints.

Abstract

This paper approximates simulation models by B-splines with a penalty on high-order finite differences of the coefficients of adjacent B-splines. The penalty prevents overfitting. The simulation output is assumed to be nonnegative. The nonnegative spline simulation metamodel is casted as a second-order cone programming model, which can be solved efficiently by modern optimization techniques. The method is implemented in MATLAB/GNU Octave.