Weighted Ensemble Vote

Problem

Weighted majority vote across models per sample. predictions[m][j] is model m's label for sample j and weights[m] is the model weight; for each sample sum the weights per label and pick the largest total.

Ties go to the smallest label. Empty predictions returns [].

Hint

Weighted voting lets stronger models dominate the ensemble decision.

Test cases

case 1

in: [[["a","b"],["b","b"],["a","a"]],[1,2,1]]

expected: ["a","b"]

case 2

in: [[[1,0,1],[0,0,1]],[3,1]]

expected: [1,0,1]

case 3

in: [[[2,1]],[0.5]]

expected: [2,1]

case 4

in: [[],[]]

expected: []

Starter codePython

def weighted_ensemble_vote(predictions, weights):
    # Your code here
    pass

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