Rejection sampling and importance sampling are two most useful methods in Bayes statistics for simulating draws from a variety of distributions. We show that rejection sampling can be viewed as a special importance sampling algorithm, and the importance sampling which uses envelope function in rejection sampling as proposal function tends to be more efficient than rejection sampling with envelope function.%拒绝性抽样法和重要性抽样法是贝叶斯统计中两种重要的产生模拟样本的方法.两种方法之间存在一些关联,拒绝性抽样法在做一些形式上的转变后可以看作一种特殊的重要性抽样法.此外,如果将拒绝性抽样法中的包络函数作为重要性抽样法中的提议密度函数,这样新形成的抽样法会比原本的拒绝性抽样法更为有效.
展开▼