A new algorithm for quadratic programming (Q1097172): Difference between revisions

We present a new finite algorithm for quadratic programming. Our algorithm is based on the solution procedures of linear programming (pivoting, Bland's rule), Hungarian methods, criss-cross method), however this method does not require the enlargement of the basis tableau as the Frank-Wolfe method does. It can be considered as a feasible point active- set method. We solve linear equation systems in order to reach an active constraint set (complementary solutions) and we solve a feasibility problem in order to check that optimality can be reached on this active set or to improve the actual solution. This algorithm is a straightforward generalization of Klafszky's and Terlaky's Hungarian method. It has nearly the same structure as Ritter's algorithm (which is based on conjugate directions), but it does not use conjugate directions.