Co-Teaching Sample Selection

Problem

Select the cleanest samples for co-teaching.

Compute the mean loss of the two models per sample, keep int(len * ratio) samples with the smallest mean loss (ties by smaller index) and return the kept indices sorted ascending.

Test cases

case 1

in: [[0.1,0.5,0.2,0.9],[0.2,0.1,0.4,0.3],0.5]

expected: [0,1]

case 2

in: [[0.4,0.1,0.3],[0.4,0.2,0.1],0.34]

expected: [1]

case 3

in: [[0.5,0.5],[0.5,0.5],1]

expected: [0,1]

Starter codePython

def coteaching_select(loss_a, loss_b, ratio):
    # Your code here
    pass

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