|
Abstract
|
In this paper, we present a new corrector-predictor interior-point method for solving semidef-
inite optimization. We use an algebraic equivalent transformation of the centering equation
of the system which defines the central path. The algebraic transformation plays an essential
role in the calculation of the new search directions. We prove that the iteration complexity
of the algorithm coincides with the best known ones for interior-point methods (IPMs). To
the best of our knowledge, this is the first corrector-predictor interior-point algorithm that
uses the search directions obtained from the desired algebraic transformation for semidefinite
optimization. Finally, some numerical experiments are provided to demonstrate the efficiency
of our new algorithm.
|