On Mon, Sep 08, 2008 at 12:12:33PM -0500, Chris Caldwell wrote: > A common mistake is to use the mean of the Poisson distribution as if it > were a separator between occurrences, rather than just a mean > difference. So they try to add something to the last occurrence, > forgetting all of the other data points. Also this mean is not a mean > from a normal distribution: long and short gaps are not symmetrically > distributed.
Hm, I thought that the negative exponential distribution (which governs the waiting time between events in a Poisson process -- the Poisson distribution represents the counts in a given interval) was "without memory" and so the distinction was unimportant? Ie. you come to the bus stop and the bus is 15 minutes away on average; you wait for five minutes with no bus, and it's _still_ 15 minutes away on average :-) /* Steinar */ -- Homepage: http://www.sesse.net/ _______________________________________________ Prime mailing list [email protected] http://hogranch.com/mailman/listinfo/prime
