On Mon, 31 Mar 2003 08:51:04 +0200, "CybercafeUser" <[EMAIL PROTECTED]> wrote:
> Dear All, > > I would like to get some refs, pointers, etc. about the use of bootstrap for > computing confidence interval of quantiles ( I need the median and > percentiles 0.01 & 0.99 in the framework of time series analysis/models). > Alhough theoretical considerations are always welcome, I would appreciate > practical implementations. I don't know how the Time series part enters in, but it must make things pretty bad. If you just wanted the CI for percentiles, you could look them up from the beta distribution. But I don't know what you are looking for, this time, when you include "time series." A google search on < nonparametric confidence> gives a number of references that (to me, at least) appear to have relevance to you question. They include the same key words -- -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
