Each week, we select a recently published Open Access article to feature. This week’s article comes from the Scandinavian Journal of Statistics and looks at classical conditional Monte Carlo methods.
The article’s abstract is given below, with the full article available to read here.
, , & (2021). Conditional Monte Carlo revisited. Scand J Statist, 1– 26. https://doi.org/10.1111/sjos.12549
Conditional Monte Carlo refers to sampling from the conditional distribution of a random vector 𝐗 given the value 𝑇(𝐗)=𝐭 for a function 𝑇(𝐗). Classical conditional Monte Carlo methods were designed for estimating conditional expectations of functions 𝜙(𝐗) by sampling from unconditional distributions obtained by certain weighting schemes. The basic ingredients were the use of importance sampling and change of variables. In the present paper we reformulate the problem by introducing an artificial parametric model in which 𝐗 is a pivotal quantity, and next representing the conditional distribution of 𝐗 given 𝑇(𝐗)=𝐭 within this new model. The approach is illustrated by several examples, including a short simulation study and an application to goodness-of-fit testing of real data. The connection to a related approach based on sufficient statistics is briefly discussed.
More Details