Title:  Highest Utility First Search: a Control Method for Multi-level
        Stochastic Design

Authors: L. Steinberg, J. Hall and B. Davison

E-Mails: lou, davison@cs.rutgers.edu

HPCD-TR-59

Abstract:

An intrinsic characteristic of stochastic optimization methods, such as 
simulated annealing, genetic algorithms and multi-start hill climbing, is
that they can be run again and again on the same inputs, each time potentially
producing a different answer.  When such algorithms are used in a design 
process with multiple levels of abstraction, where the output of one stochastic
optimizer becomes the problem statement for another stochastic optimizer,
we get an implicit tree of alternative designs.  After each optimizer run we
face a control problem of which level's optimizer to run next, and which 
design alternative to run it on.  This problem is made more difficult by the
fact that we generally can get a precise evaluation of the design alternatives
only at the lowest level (the final results), and must make do at higher levels
with only an estimate of how good a final design each alternative will lead to.