[EMAIL PROTECTED] (Henry) wrote in message news:<[EMAIL PROTECTED]>... > On 21 Jul 2002 07:41:53 -0700, [EMAIL PROTECTED] (Linda) wrote: > >Supposed that I have some data (paths) in which the birth rate and > >death rate are dependent on time in a random way i.e. for each > >time-instant, the birth & death rates are some random variable. I have > >tried to plot the histogram of them and they both can be approximate > >by an exponential PDF, respectively. My aim is to model the birth & > >death of these paths for certain period of time. Since simple B-D > >process only allows a homogenous birth & death rate, I wonder what is > >the more appropriate method that I can use to model my problem. > > Is there some sort of time dependence or corelation? If the birth and > death rates at any instant are independent of those at any other time, > I would have guessed that you would get the same result as a simple > B&D process using the means of the random variables.
Do you mean that I need to check the correlation between the birth & death rates? I have looked at the general pattern of the birth and death rates and it seems that they are proportional to each other. What do you mean by means of random variables?? Thanks. Linda . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
