Manjeet Dahiya

Probability Distribution

This post presents the definition of probability distribution and cumulative distribution.

Distribution of a Random Variable

We have probabilities of the different outcomes of an experiment, consequently, we get the probabilities of the different values that an associated random variable takes. The distribution of a random variable ($X$) is a collection of probabilities of events of form ${X \in C}$, for all possible subsets $C$ of real numbers. In simpler terms, the distribution of a random variable gives the probabilities of all possible values (including ranges) taken by a random variable.

A discrete distribution is defined in terms of probability function (pf), which also called probability mass function. For continuous random variables, distribution is defined in terms of probability density function (pdf). Thus, a pf or pdf helps define the distribution of a random variable. A summation or integration can be used to compute the same.

Note: $Pr(X = c)$ is $0$ for any $c$ in case of a continuous random variable. The reason is that the area/integration is zero, another explanation is that the continuos random variable takes infinite number of values, the probability of taking a particular value is zero. However, the probability of ranges would turn out to be non-zero because of integration over a bigger range.

All pf and pdf must satisfy following conditions to obey the axioms of probability:

Cumulative Distribution Function

A cumulative distribution function $F(x)$ is defined as:

$F(x) = Pr( X \le x)$ for $x \in (-\infty, \infty)$

Note that the same definition is applicable for discrete as well as continuous random variables. Unlike before, we do not need two different definitions for discrete and continuous.

Properties of $F(x)$:

Examples

Let us define the probability distribution of a random variable ($X$) that represents the number of heads on tossing two fair coins. The random variable is

Its probability distribution can be written as:

$Pr(X = 0) = 1/4$ (event: {TT})

$Pr(X = 1) = 1/2$ (event: {HT, TH})

$Pr(X = 2) = 1/4$ (event: {HH})

Point to note is that probability distribution is nothing but a collection of probabilities of the different values taken by the respective random variable.


© 2018-19 Manjeet Dahiya