Speaker:   Jie Sun
  Dept. of Decision Sciences
  National University of Singapore


Title: Semismooth Matrix Equations

Solving a semidefinite complementarity problem (SDCP) is equivalent to finding a root of a certain matrix-valued nonsmooth equation. A smoothing Newton method can be used to solve the nonsmooth matrix equation derived from the SDCP. The squared smoothing function is shown to be strongly semismooth. This property allows us to establish quadratic convergence results without restrictive assumptions.