Direct search methods for nonsmooth optimization

We present analytical tools that enable development of globally convergent direct search methods for minimization of nonsmooth locally Lipschitz functions. The theory is illustrated with the design of new algorithms.