Resolution of general linear systems: numerical methods and applications

Abstract

[en] The Singular Value Decomposition and the complete orthogonal decomposition algorithms are two numerical methods based on orthogonal matrix transformations. Among other usages, both are used for least square approximately solving general linear systems. Not all the systems have a unique solution, but all the possible solutions can be found. Besides, in this project, these two methods are going to be compared each other and then see how apply them to nonlinear systems in a geodesic scope.