A Friendly Smoothed Analysis of the Simplex Method: Gaussian-Perturbed Shadows Have at Most 3 + 128πe²d²√(log n)σ⁻²(1 + 4σ√(d log n))(1 + 16σ√(log n)) Expected EdgesResearch Paper
The simplex method solves linear programs quickly in practice, yet for most pivot rules there are inputs on which it takes exponentially many steps (Klee and Minty, 1972). Smoothed analysis, introduced by Spielman and Teng (JACM 2004), explains the gap: it measures the expected running time when every constraint row of a worst-case input is perturbed by a…