Task 1

In the context of the k-arm bandit problem, what is the primary challenge faced by the decision-making algorithm?

  1. Determining the optimal sequence of actions before the first time-frame begins.
  2. Balancing exploration of unknown actions with exploitation of known high-reward actions.
  3. Predicting how the reward for each action will change over time.
  4. Minimizing the total number of actions taken within the time period T.

[Total 3 marks]