Issue No. 02 - March/April (2000 vol. 2)
In recent years, several new Monte Carlo methods have proven to be very effective for sampling from multimodal energy landscapes, like those found near a first-order phase transition or in a glassy material. In this column, we will summarize the theoretical structure of one of these methods, the multicanonical method,1,2 as it is perhaps the most enigmatic of the new algorithms. Special emphasis will be on the manner by which it is an importance-sampling method.
Naomichi Hatano, Jim Gubernatis, "The Multicanonical Monte Carlo Method", Computing in Science & Engineering, vol. 2, no. , pp. 95-102, March/April 2000, doi:10.1109/MCSE.2000.10007