Model Confidence Set Trim

Problem

Trim a set of models to a confidence set. Compute each model's mean loss, then return the indices of the `keep` models with the smallest means, sorted in ascending index order.

Return an empty list for empty input or keep <= 0.

Hint

Ties are broken by original index because the sort key includes it.

Test cases

case 1

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

expected: [0,1]

case 2

in: [[[5],[1],[3]],1]

expected: [1]

case 3

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

expected: [0,1]

case 4

in: [[],1]

expected: []

Starter codePython

def model_confidence_set_trim(losses, keep):
    # Your code here
    pass

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