Uncertainty Penalty Coefficient

Problem

Penalize optimistic Q values by uncertainty: return mean_q[a] - beta * std_q[a] for each action.

Hint

Subtracting beta times the standard deviation yields pessimistic value estimates.

case 1

in: [[5,4],[1,2],0.5]

expected: [4.5,3]

case 2

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

expected: [-3]

case 3

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

expected: [1,2]

Starter codePython

def uncertainty_penalty(mean_q, std_q, beta):
    # Your code here
    pass

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