Limit theorems for weighted samples with applications to sequential Monte Carlo methods (Q955144)

In the last decade, sequential Monte Carlo (SMC) methods emerged as a key tool in computational statistics. These algorithms approximate a sequence of distributions by a sequence of weighted empirical measures associated to a weighted population of particles, which are generated recursively. In this paper the large sample properties of weighted particle approximations are studied, when the number of particles tends to infinity. The particles are not independent implying difficulties in the analysis. The paper establishes both the law of large numbers and central limit theorems under minimal conditions. The results apply to schemes such as resampling and move algorithms or auxiliary particle filters, and other schemes.