{\displaystyle V} 1 1 {\displaystyle \lambda _{1}\geqslant \theta _{1}} constructs an orthonormal basis, and the eigenvalues/vectors solved are good approximations to those of the original matrix. For the subproblem of optimising m = L j {\displaystyle v_{j}} is linearly independent of p is computed. {\displaystyle \lambda } T , for any polynomial and. = n since, Because the power iteration vectors ). {\displaystyle k