Thursday, September 1st, 2005
Quasi-Monte Carlo Metropolis algorithm
According to the PNAS Journal (a favorite of Berkely Groks), the quasi-Monte Carlo Metropolis algorithm can get your results much quicker if your MC problem happens to fit their conditions: it has to be “completely uniformly distributed” (CUD). If you have a CUD problem, PNAS can solve it (those of your giggling at this are immature).
Thanks to my friend Sotos for pointing this one out.
