Newton's method and secant methods: a longstanding relationship from vectors to matrices (Q658578): Difference between revisions

Let us consider the following nonlinear problem: for a given \(F:H\rightarrow H\), find \(X_{*}\) such that \(F(X_{*})=0\), where \(H\) is a normed space, and \(F\) is a Fréchet differentiable map. In case of the space of square matrices, the problem can be rewritten as follows. For a given \(F:\mathbb{C}^{n\times n}\rightarrow \mathbb{C}^{n\times n}\), find \(X_{*}\in\mathbb{C}^{n\times n}\) such that \(F(X_{*})=0\). Such nonlinear matrix problems play an important role in different scientific topics such as applied statistics, control theory, and financial mathematics. The aim of this review paper is to present and discuss the state of the art in the use of Newton and secant methods for solving such nonlinear problems. Especially, the authors present a unified approach for describing the features of the classic methods for solving the problems of finding matrix inverse, where \(F(X)=X^{-1}-A\), matrix \(p\)-th root, where \(F(X)=X^{p}-A\), \(p\in\mathbb{N}\), \(p\geq 2\), sign of a matrix, where \(S=\mathrm{sign}(A)=A(A^{2})^{-1/2}\), Riccati equations \(F(X)=XA+A^TX-XBB^TX+C^TC\), quadratic equations \(AX^2+BX+C=0\), matrix rational equations, matrix logarithm and matrix exponential. A very long (179 items) list of valuable references is also given.