Efficient algorithms for solving the \(p\)-Laplacian in polynomial time (Q2209522)

The author provides new numerical algorithms, based on the barrier method of convex optimization in order to solve a non homogeneous Dirichlet boundary value problem attached to the $p$-Laplace equation. He shows that his algorithms compute a solution to $p$-Laplacian, with the exponent $p$ larger or equal with 1, to any given tolerance in polynomial time, using $O(\sqrt{n\log n})$ Newton iterations, and an adaptive stepping variant converges in $O(\sqrt{n\log 2n})$ Newton iterations. Some numerical experiments are carried out in order to confirm these scalings.