25 April 2022 0 9K Report

How can we convert L2 error function or mean squared error function to give different weights to underestimation and overestimation ?

Or does there any error function exist which is differentiable at all points, has varying gradient for different values of x, and at the same gives different weightage to overestimation and underestimation ?

More Titas De's questions See All
Similar questions and discussions