I read Iterative methods for optimization by C. Kelley (PDF) and I'm struggling to understand proof of
Notes on notation: S is a simplex with vertices x_1 to x_{N+1} (order matters), some edges v_j = x_j - x_1 that make matrix V = (v_1, \dots, v_n) and \sigma_+(S) = \max_j \lVert...