A new algorithm for the linear complementarity problem allowing for an arbitrary starting point

The linear complementarity problem is to find nonnegative vectors which are affinely related and complementary. In this paper we propose a new complementary pivoting algorithm for solving the linear complementarity problem as a more efficient alternative to the algorithms proposed by Lemke and by Talman and Van der Heyden. The algorithm can start at an arbitrary nonnegative vector and converges under the same conditions as Lemke's algorithm.

Kremers, Johannes and Dolf Talman (1994), A new algorithm for the linear complementarity problem allowing for an arbitrary starting point, Mathematical Programming 63 , 235-252