And error measures involving sum of absolute values (i.e., L1 norm) are central to methods like lasso (https://en.wikipedia.org/wiki/Lasso_(statistics) https://

And error measures involving sum of absolute values (i.e., L1 norm) are central to methods like lasso (https://en.wikipedia.org/wiki/Lasso_(statistics) https://en.wikipedia.org/wiki/Lasso_(statistics)) and their cousins.