CEM Mean Update

Problem

Update the cross-entropy method distribution by averaging the elite samples.

Take the top max(1, int(n * elite_frac)) samples by score (ties broken by smallest index) and return their mean vector. No samples returns [].

Hint

Only the elite fraction of samples shapes the next search distribution.

case 1

in: [[[0,0],[1,2],[2,4],[3,6]],[1,4,3,2],0.5]

expected: [1.5,3]

case 2

in: [[[1],[2],[3]],[5,5,1],1]

expected: [2]

case 3

in: [[[0],[1]],[1,2],0.6]

expected: [1]

Starter codePython

def cem_mean_update(samples, scores, elite_frac):
    # Your code here
    pass

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