Exponential Distribution
(see Piecewise Functions)
For
left=-1; right=5;
bottom=-1;top=2;
---
y = e^{-x} \{ x \ge 0 \}
y = 0 \{ x < 0 \}
Support and CDF
(see integral)
For
left=-1; right=5;
bottom=-1;top=2;
---
y = 1 - e^{-x} \{ x \ge 0 \}
y = 0 \{ x < 0 \}
Alternate notation:
Expectation and Variance
(expected value and variance)
Given
Let
Let
Let
Explanation for (1): Since
Explanation for (2): from the proof of expected value, we found that
Explanation for (3): see LOTUS
Memorylessness
The exponential distribution is memoryless
We want to prove
proof
- Waiting time interpretation
- Detector for radioactive decay
- Waiting for detector to go off has same probability if you came in later
- Detector for radioactive decay
Relationships to Other Distributions
- Bernoulli Distribution
- Sequence of 1 and 0
- "Waiting time" for first success is the Geometric Distribution
- Poisson Distribution
- Poisson process: continuous version of Bernoulli
- "Waiting time" for first success is the exponential distribution
- Let
be the number of arrivals in a time interval of length of a Poisson process with intensity - Let
be the waiting time for the first arrival
then
proof
Let CDF of
: probability you had to wait more than for first success : this is Poisson, which measures the number of successes in some interval, so
todo what??!??
Examples
Suppose
solution
Median: 50th percentile.
Let
Note: "percentile" refers to an actual
[!example]
City
- What is the expected waiting time?
- Find
solution
Let
- Should wait on average
minutes
[!example]
City
- What is the expected waiting time?
- Find
solution
Let
- ???