Mean Squared Logarithmic Error

Problem

Compute the mean squared logarithmic error: mean((log1p(y_i) - log1p(pred_i))^2).

Targets and predictions are assumed non-negative. Empty input returns 0.0.

Hint

Apply log1p to both sides before squaring, so ratios matter more than absolute gaps.

case 1

in: [[1,2,3],[1,2,3]]

expected: 0

case 2

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

expected: 0.4804530139182014

case 3

in: [[1,3],[2,1]]

expected: 0.32242748390568343

case 4

in: [[],[]]

expected: 0

Starter codePython

import math
def mean_squared_log_error(y_true, y_pred):
    # Your code here
    pass

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