CORAL Covariance Align

Problem

Compute the CORAL domain-alignment loss.

Build the population covariance matrix of each feature set (rows are samples), then return ||C_source - C_target||_F^2 / (4 * d^2) rounded to 4 decimal places, where d is the feature dimension.

Test cases

case 1

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

expected: 0

case 2

in: [[[1],[2]],[[1],[4]]]

expected: 1

case 3

in: [[[1],[2],[3]],[[2],[4],[6]]]

expected: 1

Starter codePython

def coral_loss(source, target):
    # Your code here
    pass

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