Blurb::  Use the multilevel MCMC algorithm. 

Description::  
Selects the multilevel algorithm described in \cite Prudencio.
 
<b> Default Behavior </b>

Five MCMC algorithm variants are supported: \c dram, \c
delayed_rejection, \c adaptive_metropolis, \c metropolis_hastings, and
\c multilevel.  The default is \c dram.

<b> Usage Tips </b>

The multilevel algorithm is a more experimental algorithm than the
other MCMC approaches mentioned above.  It works well in cases where
the prior can be "evolved" to a posterior in a structured way.
Currently, the multilevel option is not in production form. 

Topics::	bayesian_calibration	
Examples::
\verbatim
method,
        bayes_calibration queso
          multilevel
          samples = 10000 seed = 348
\endverbatim

Theory::
Faq::
See_Also::	
