An Improved Infeasible Full Nesterov-Todd Interior-Point Algorithm for the Linear Complementarity Problem over Symmetric Cones

In this talk an infeasible full Nesterov-Todd step interior-point method for Linear Complementarity Problems over Symmetric Cones is considered. Using several new results from Euclidean Jordan algebras and associated symmetric cones, a sharper quadratic convergence result than previously known is established, leading to a wider quadratic convergence neighborhood of the central path for the feasibility steps of the algorithm. However, the best iteration bounds known for the infeasible short-step methods, is still achieved.