Beschreibung:
<jats:p>
We show how to solve the maximal cut and partition problems using a randomized algorithm based on the
<jats:italic>cross-entropy</jats:italic>
method. For the maximal cut problem, the proposed algorithm employs an auxiliary Bernoulli distribution, which transforms the original deterministic network into an associated stochastic one, called the
<jats:italic>associated stochastic network</jats:italic>
(ASN). Each iteration of the randomized algorithm for the ASN involves the following two phases:(1) Generation of random cuts using a multidimensional
<jats:italic>Ber</jats:italic>
(
<jats:bold>p</jats:bold>
) distribution and calculation of the associated cut lengths (objective functions) and some related quantities, such as rare-event probabilities.(2) Updating the parameter vector
<jats:bold>p</jats:bold>
on the basis of the data collected in the first phase.We show that the
<jats:italic>Ber</jats:italic>
(
<jats:bold>p</jats:bold>
) distribution converges in distribution to a degenerated one,
<jats:italic>Ber</jats:italic>
(
<jats:bold>p</jats:bold>
<jats:sub>
<jats:italic>d</jats:italic>
</jats:sub>
<jats:sup>*</jats:sup>
),
<jats:bold>p</jats:bold>
<jats:sub>
<jats:italic>d</jats:italic>
</jats:sub>
<jats:sup>*</jats:sup>
= (
<jats:italic>pd</jats:italic>
,1,...,
<jats:italic>pd,n</jats:italic>
) in the sense that someelements of
<jats:bold>p</jats:bold>
<jats:sub>
<jats:italic>d</jats:italic>
</jats:sub>
<jats:sup>*</jats:sup>
, will be unities and the rest zeros. The unity elements of
<jats:bold>p</jats:bold>
<jats:sub>
<jats:italic>d</jats:italic>
</jats:sub>
<jats:sup>*</jats:sup>
uniquely define a cut which will be taken as the estimate of the maximal cut. A similar approach is used for the partition problem. Supporting numerical results are given as well. Our numerical studies suggest that for the maximal cut and partition problems the proposed algorithm typically has polynomial complexity in the size of the network.
</jats:p>