Sequential Monte Carlo Methods in Practice

Monte Carlo methods are revolutionising the on-line analysis of datain fields as diverse as financial modelling, target tracking andcomputer vision. These methods, appearing under the names of bootstrapfilters, condensation, optimal Monte Carlo filters, particle filtersand survial of the fittest, have made it possible to solve numericallymany complex, non-standarard problems that were previouslyintractable.This book presents the first comprehensive treatment of thesetechniques, including convergence results and applications totracking, guidance, automated target recognition, aircraft navigation,robot navigation, econometrics, financial modelling, neuralnetworks,optimal control, optimal filtering, communications,reinforcement learning, signal enhancement, model averaging andselection, computer vision, semiconductor design, population biology,dynamic Bayesian networks, and time series analysis. This will be ofgreat value to students, researchers and practicioners, who have somebasic knowledge of probability.Arnaud Doucet received the Ph. D. degree from the University of Paris-XI Orsay in 1997. From 1998 to 2000, he conducted research at theSignal Processing Group of Cambridge University, UK. He is currentlyan assistant professor at the Department of Electrical Engineering ofMelbourne University, Australia. His research interests includeBayesian statistics, dynamic models and Monte Carlo methods.Nando de Freitas obtained a Ph.D. degree in information engineeringfrom Cambridge University in 1999. He is presently a researchassociate with the artificial intelligence group of the University ofCalifornia at Berkeley. His main research interests are in Bayesianstatistics and the application of on-line and batch Monte Carlomethods to machine learning.