In this paper we generalize the primal--dual cone affine scaling algorithm of Sturm and Zhang to semidefinite programming. We show in this paper that the underlying ideas of the cone affine scaling algorithm can be naturely applied to semidefinite programming, resulting in a new algorithm. Compared to other primal--dual affine scaling algorithms for semidefinite programming, our algorithm enjoys the lowest computational complexity.

, ,
hdl.handle.net/1765/1397
Econometric Institute Research Papers
Erasmus School of Economics

Berkelaar, A., Sturm, J. F., & Zhang, S. (1996). Polynomial Primal-Dual Cone Affine Scaling for Semidefinite Programming (No. EI 9667-/A). Econometric Institute Research Papers. Retrieved from http://hdl.handle.net/1765/1397