Target Encoding Leakage Check

Problem

Flag a target-encoding leakage risk: return True when the fraction of unique categories exceeds threshold.

Empty input returns False.

Hint

High-cardinality features memorized by target encoding can leak labels.

case 1

in: [["a","a","b","b"],0.4]

expected: true

case 2

in: [["a","a","b","b"],0.6]

expected: false

case 3

in: [["a","a","b","b"],0.5]

expected: false

case 4

in: [[],0.4]

expected: false

Starter codePython

def target_encoding_leakage(categories, threshold):
    # Your code here
    pass

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