Distance-Weighted KNN Vote

Problem

Predict a label by distance-weighted k-nearest-neighbor voting using Euclidean distance. Weight each neighbor by 1/d; if a zero distance is found, immediately return that neighbor's label.

Ties are broken by the smallest label. If there are no neighbors to vote, return None.

Hint

Nearer neighbors get larger 1/d weights; an exact match short-circuits the vote.

Test cases

case 1

in: [[[0],[1],[3]],["a","b","b"],[0.5],3]

expected: "b"

case 2

in: [[[1],[2]],[5,9],[1],2]

expected: 5

case 3

in: [[[0],[10]],[1,2],[1],1]

expected: 1

case 4

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

expected: null

Starter codePython

def knn_weighted_vote(X_train, y_train, x_test, k):
    # Your code here
    pass

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