- (iii) Of course, it is obvious from (NI.16) that Monte Carlo integration gets better as standard deviation gets smaller.
- Using definition of (or remembering its name ``variance''), it follows that method is best when is slowly varying over domain.
- For instance, if constant, and method is exact.
- Importance sampling (``sampling where integrand important'') is a technique for reducing variation of .
- Thus if
- We now apply Monte Carlo method generating
**y**uniformly.

Geoffrey Fox, Northeast Parallel Architectures Center at Syracuse University, gcf@npac.syr.edu