An Introduction to Sequential Monte Carlo Methods

摘要: Many real-world data analysis tasks involve estimating unknown quantities from some given observations. In most of these applications, prior knowledge about the phenomenon being modelled is available. This allows us to formulate Bayesian models, that distributions for and likelihood functions relating Within this setting, all inference on based posterior distribution obtained Bayes’ theorem. Often, observations arrive sequentially in time one interested performing on-line. It therefore necessary update as become Examples include tracking an aircraft using radar measurements, a digital communications signal noisy or volatility financial instruments stock market data. Computational simplicity form not having store might also be additional motivating factor sequential methods.