A wide neighborhood arc-search interior-point algorithm for convex quadratic programming with box constraints and linear constraints

In this paper, a wide neighborhood arc-search interior-point algorithm for convex quadratic programming with box constrains and linear constraints (BLCQP) is presented. The algorithm searches the optimizers along the ellipses that approximate the entire central path. Assuming a strictly feasible initial point is available, we show that the algorithm has O(n(3/4) log (x(0)-l)(T) s(0)+(w-x(0))(T) t(0)/epsilon) iteration complexity bound, which is the best known complexity result for such methods. The numerical results show that our algorithm is effective and promising.

Automated summary

This paper presents a wide neighborhood arc-search interior-point algorithm for convex quadratic programming with box constrains and linear constraints (BLCQP). The algorithm searches for optimizers along ellipses approximating the central path. With a strictly feasible initial point, the algorithm has the best known complexity result for such methods, as shown by numerical results demonstrating its effectiveness and promise.