Stochastic Approximation Monte Carlo: Theory and Applications 


Abstract: 	

The stochastic approximation Monte Carlo (SAMC) algorithm has been 
recently proposed by F. Liang and his co-authors as a general Monte 
Carlo and optimization algorithm. 
A remarkable feature of the algorithm is that it avoids the 
local-trap problem suffered by conventional MCMC algorithms; 
 the algorithm can self-adjust its acceptance rate of local 
Metropolis-Hastings moves such that each subregion of the sample 
space can be sampled with a desired frequency. In this talk, I will 
first provide an overview of the algorithm from both the theoretical 
and practical aspects. I will then discuss applications of the algorithm 
to a variety of statistical and bioinformatical problems, including 
neural network training, disease mapping, phylogenetic tree reconstruction,
among others.