Solving Systems of Nonlinear Equations using Hybrid Optimization Methods

Abstract

This licentiate thesis considers various hybrid methods for solving systems of nonlinear equations. Extensive numerical experiments confirm a research hypothesis that one particular form of hybridization of the conjugate gradient algorithm with Newton’s method outperforms other hybrid strategies considered in the thesis. Although the findings are preliminary and the computational experiments have certain limitations, this study offers solid insights by highlighting the most promising hybridization techniques for further research and potential advancements in solving nonlinear systems.