Learning agents have to deal with the exploration-exploitation dilemma.

This paper proposes an algorithm to better solve the exploration-exploitation dilemma faced by model-less reinforcement learning agents. The main contribution is twofold: (1) The two facets of the exploration-exploitation dilemma are distinguished: in some cases, the agent faces a non

been searching for the optimal solution to this dilemma. Here we present a novel model in which the exploration strategy itself is dynamic and varies with time in order to optimize a definite goal, such as the acquisition of energy, money, or prestige. Our model produced four very distinct phases

author versus an author whose books vary widely in quality. All of these decisions involve balancing the conflicting demands of exploiting previous knowledge in order maximize payoffs versus exploring less-known options in order to gain information about the currently optimal course of action. Indeed

The exploration-exploitation dilemma is a recurrent adaptive problem for humans as well as non-human animals. Given a fixed time/energy budget, every individual faces a fundamental trade-off between exploring for better resources and exploiting known resources to optimize overall performance under

and dynamic environments

Full instructions for the task Before taking part in the experiment, participants read a set of illustrated onscreen instructions describing the task and its mechanics. Here we provide the full text for these instructions. Each bullet point corresponds to a single screen in the instructions. In the interests of space we have not included the illustrations themselves. • Welcome! Thank you for volunteering for this experiment. • In this experiment we would like you to choose between two one-armed bandits of the sort you might find in a casino. • The one-armed bandits will be represented like this • Every time you choose to play a particular bandit, the lever will be pulled like this... •... and the payoff will be shown like this. For example, in this case, the left bandit has been played and is paying out 77 points. • Each bandit tends to pay out about the same amount of reward on average, but there is variability in the reward on any given play. • For example, the average reward for the bandit on the right might be 50 points, but on the

Abstract. Reinforcement learning policies face the exploration versus exploitation dilemma, i.e. the search for a balance between exploring the environment to find profitable actions while taking the empirically best action as often as possible. A popular measure of a policy’s success in addressing

of reinforcement learning, including trading off exploration and exploitation, establishing the foundations of the field via Markov decision theory, learning from delayed reinforcement, constructing empirical models to accelerate learning, making use of generalization and hierarchy, and coping with hidden state

Tabu Search is a metaheuristic that guides a local heuristic search procedure to explore the solution space beyond local optimality. One of the main components of tabu search is its use of adaptive memory, which creates a more flexible search behavior. Memory based strategies are therefore