Furthermore, the computational complexity of the policy-iteration and simplex methods is superior to that of the only known strongly polynomialtime interior-point algorithm [Ye, Y. 2005.