A convex quadratic programming based on a novel type of parameterized kernel function

Abstract

This study introduces primal-dual interior-point methods  for convex quadratic programming, based on a novel parameterized hyperbolic kernel function. Under certain conditions and by using simple analytical approaches, along with a specific choice of a key parameter, the proposed method achieves one of the best known iteration efficiencies for large-update methods. Numerical experiments are also presented to demonstrate the practical effectiveness of the approach.