Results on Newton methods. I: A unified approach for constructing perturbed Newton-like methods in Banach space and their applications (Q5917794): Difference between revisions

The author considers the problem of approximating a locally unique solution \(x^*\) of the equation \(F(x)= 0\), where \(F\) is a nonlinear operator defined on some convex subset of a Banach space \(E_1\) with values in a Banach space \(E_2\). The author introduces the Newton-like method by \[ y_n= x_n- F'(x_n)^{- 1} F(x_n),\quad x_{n+ 1}= y_n- z_n, \] where \(F'(x_n)\) is the Fréchet derivative of \(F\) at \(x_n\), and \(z_n\in E_1\) are points to be determined. Under certain assumptions of \(F\) the sequence \(\{x_n\}\) generated by the above algorithm converges to a locally unique solution \(x^*\). The order of convergence for the iteration is also examined. Under special choices of the points \(z_n\), \(n\geq 0\), his results can be reduced to the ones in previous papers. The monotone convergence of the method is also examined in the paper. Moreover, the author shows that under weaker hypotheses his iterations converge faster than others already studied in the literature. [For part II see ibid. 74, No. 2-3, 143-159 (1996; reviewed below)].