This preview shows page 1. Sign up to view the full content.
Unformatted text preview: d geometric distributions by showing that the
exponential distribution is the limit of scaled geometric distributions. In essence, the exponential 80 CHAPTER 3. CONTINUOUS-TYPE RANDOM VARIABLES distribution is the continuous time analog of the geometric distribution. When systems are modeled
or simulated, it is useful to be able to approximate continuous variables by discrete ones, because
digital computers can only represent quantities to ﬁnite precision.
Let λ > 0 and let h be a small positive number. We shall deﬁne a discrete-type random variable
T which is approximately exponentially distributed with parameter λ. Think of T as the lifetime
of a lightbulb. Use time clicks of duration h. Let L be the number of clicks until the lightbulb
fails. Suppose L is geometrically distributed, with parameter p = hλ. Thus, the pmf of L is given
by pL (k ) = (1 − hλ)k−1 hλ, and the interpretation is that the lightbulb fails during each click with
probability hλ, given that it didn’t fail previously. So (1 − hλ)k−1 is the probabi...
View Full Document
This note was uploaded on 02/09/2014 for the course ISYE 2027 taught by Professor Zahrn during the Spring '08 term at Georgia Institute of Technology.
- Spring '08
- The Land