Revenue Optimization Threshold

Problem

Choose the probability threshold that maximizes expected profit. For each unique probability t, act on samples with p >= t and earn p * value - cost per action.

Return [best_threshold, best_profit]; ties keep the smallest threshold and empty input returns [0.0, 0.0].

Hint

Act whenever the expected value p * value exceeds the action cost.

Test cases

case 1

in: [[0.9,0.6,0.3],[10,10,10],2]

expected: [0.3,12]

case 2

in: [[0.2,0.25],[100,100],10]

expected: [0.2,25]

case 3

in: [[0.1],[100],5]

expected: [0.1,5]

case 4

in: [[0.01],[1],10]

expected: [0.01,-9.99]

Starter codePython

def revenue_optimization(probs, values, cost):
    # Returns [best_threshold, best_profit]
    # Your code here
    pass

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