The limitations of standard algorithmic design aspects are being investigated. Assertions that certain classes of algorithms are too weak to solve certain computational problems are being established. It is an important aspect of this research that the classes of algorithms being investigated are not defined in the usual way by resource limitations (such as running time being bounded by a polynomial function of input length) but by combinational properties aimed at capturing intuitive notions of different algorithm design techniques (such as divide-and-conquer, local optimization, and others). The P.I. is a leader in the proposed approach to dealing with problems that lie near the boundary between the tractable and intractable computational problems. Work of this sort has the potential for effecting the practice with which important practical problems are dealt with.
