It's one of the most widely used measures of inequality. It measures the inequality of a distribution using a Lorenz curve. A Lorenz curve is a cumulative frequency curve that compares the distribution of a specific variable with a uniform distribution that represents the equality. The value of a Gini coefficient ranges from 0 to 1, where 0 represents perfect equality and the value of 1 perfects inequality. It's basically half of the relative absolute mean difference.

So, in our meta learning setting, we can calculate the Gini coefficient as follows: