by IanRayner | Oct 18, 2010 |
A constant source of nagging doubt in system design is the use of parametric probability density functions. For the most part it is better to use empirical distributions i.e. use the data itself to determine the liklihood of a given future occurrance. Of course, this...
