Section 8.1 Time vs Changes
With Bernoulli-based distributions, you considered the number of successes in a fixed number of trials (Binomial) or the number of trials until obtaining a fixed number of successes (Negative Binomial/Geometric). For each of these, everything was discrete--trials and successes. Another option is to investigate the relationship between number of successes (or "changes") relative to some interval of time. That is, the number of trials for Bernoulli-based distributions is replace with an interval of time. This approach involves the discrete number of changes with a continuous interval of time.
Definition 8.1.1. Poisson Process.
Consider an experiment that takes place over some interval of time. A Poisson process is a course of action in which:
Successes in non-overlapping subintervals are independent of each other.
The probability of exactly one success in a sufficiently small interval of length h is proportional to h. In notation, P(one success) = \(\lambda h\text{.}\)
The probability of two or more successes in a sufficiently small interval is essentially 0.
These assumptions lead to the following distributions discussed in this chapter.
In this chapter, you will investigate the following distributions:
Poisson - the interval is fixed and X measures the variable number of successes.
Exponential - the number of successes is fixed--at 1--and X measures the variable interval length needed to get that success.
Gamma - the number of successes is fixed and X measures the variable interval needed to get the desired number of successes.
