Pareto distribution

80% of the outcomes are due to 20% of causes, only Pareto distributions with shape value $\alpha=\log_4 5=1.16$ reflect this.

in machine learning, we can do feature ablation based on its Pareto distribution.

definition

if $X$ is a random variable with Pareto distribution (Type I), then the survival function is given by:

$$\overline{F(x)} = Pr(X > x) = \begin{cases} \displaystyle \left(\frac{x_m}{x}\right)^\alpha, & x \ge x_m,\\[1em] 1, & x < x_m. \end{cases}$$

where $x_m$ is the (necessarily positive) minimum possible value of $X$, and $\alpha$ is a positive parameter.