Lecture 2 - Omar GHATTAS (U. Texas, Austin): The MAP estimate: large-scale optimization via inexact Newton-CG, adjoint-based gradients and Hessians via the Lagrangian formalism. - Amphi Darboux