Bernoulli Distribution
Definition1
For $p \in [0, 1]$, a discrete probability distribution with the following probability mass function is referred to as a Bernoulli distribution.
$$ f(x) = p^{x}(1-p)^{1-x}, \qquad x = 0, 1 $$
Description
This distribution is used when describing an experiment with only two possible outcomes, such as a coin toss. Because there are two possible outcomes, $x = 1$ is commonly referred to as a success and $x = 0$ as a failure. The probability of success is denoted as $p$, and the probability of failure as $q = 1 - p$. Conducting an experiment where there are only two possible outcomes is known as a Bernoulli trial.
When the number of trials is generalized to $n$ times, it results in a binomial distribution. Conversely, a Bernoulli distribution can be viewed as a special case of the binomial distribution when $n = 1$ becomes $\operatorname{Bin}(1, p)$.
When the possible outcomes (categories) are generalized from two to $k$, it becomes a categorical distribution. If both the number of trials and categories are generalized, it becomes a multinomial distribution.
Category
Number of Trials | $1$ times | $n$ times |
$2$ categories | Bernoulli distribution | Binomial distribution |
$k$ categories | Categorical distribution | Multinomial distribution |
Basic Properties
Moment Generating Function
The moment generating function of a Bernoulli distribution is as follows.
$$ m(t) = 1 - p + pe^{t} = q + pe^{t}, \qquad q = 1 - p $$
Mean and Variance
If $X \sim \operatorname{Bin}(1, p)$, then
$$ E(X) = p $$ $$ \Var(X) = p(1-p) = pq, \qquad q = 1 - p $$
Hogg et al. (2018). Introduction to Mathematical Statistcs(8th Edition): p155-157 ↩︎