Performance Decay Monitor

Problem

Monitor model decay: compare the mean error of the last window against the mean error of the first window times threshold. Return True when recent > baseline * threshold.

Return False when errors is shorter than window or window is not positive. If the baseline is non-positive, return True only when the recent mean is positive.

Hint

A threshold of 2 flags when recent error has doubled versus the baseline.

Test cases

case 1

in: [[0.1,0.1,0.1,0.3,0.3,0.3],3,2]

expected: true

case 2

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

expected: false

case 3

in: [[1,2],1,1.5]

expected: true

case 4

in: [[],3,2]

expected: false

Starter codePython

def performance_decay_monitor(errors, window, threshold):
    # Your code here
    pass

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