A type of modified BFGS algorithm with any rank defects and the local \(Q\)-superlinear convergence properties (Q854430)

The authors present a modified BFGS algorithm for solving unconstrained optimization problems of the form \[ \text{minimize}\,f(x)\mid x\in\mathbb R^n. \] The main idea of the algorithm is to add a modified term to the convex function in order to obtain an equivalent model, then simplify the model to get the modified BFGS algorithm. The superlinear convergence of the algorithm is proved. The given algorithm is more efficient for solving singular unconstrained optimization problems in computing amount and complication. Numerical examples are given.