Bayesian Regret Lite

Problem

Compute the cumulative pseudo-regret of a bandit algorithm.

For each arm add counts[a] * (max(means) - means[a]), the expected loss from pulling suboptimal arms. Return the total.

Hint

Only pulls of suboptimal arms contribute to regret.

case 1

in: [[0.2,0.5],[10,5]]

expected: 3

case 2

in: [[0.1,0.4,0.6],[3,4,5]]

expected: 2.3

case 3

in: [[1],[7]]

expected: 0

Starter codePython

def bayesian_regret_lite(means, counts):
    # Your code here
    pass

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