MSE Gradient Per Element

Problem

For mean squared error L = (1/n) * sum_i (pred_i - target_i)^2, the gradient with respect to each prediction is 2 * (pred_i - target_i) / n.

Given predictions and targets, return the gradient list.

Hint

The 1/n from the mean scales every element.

case 1

in: [[1,2],[0,0]]

expected: [1,2]

case 2

in: [[1,1,1],[1,1,1]]

expected: [0,0,0]

case 3

in: [[0,3],[1,1]]

expected: [-1,2]

Starter codePython

def mse_gradient_per_element(preds, targets):
    # Your code here
    pass

3 test cases · run your code against them in the editor