We describe how a standard reinforcement learning algorithm can be changed to include a
second contextual input that is used to modulate the learning in the original algorithm. The …

This paper presents a system that transfers the results of prior learning to speed up
reinforcement learning in a changing world. Often, even when the change to the world is …

When applying reinforcement learning in domains with very large or continuous state
spaces, the experience obtained by the learning agent in the interaction with the …

Much research in Reinforcement Learning (RL) has focused on learning algorithms and
generalization using simple representation languages for states and actions. Recently, there …

Reinforcement learning (RL) is an algorithmic theory for learning by experience optimal
action control. Two widely discussed problems within this field are the temporal credit …

In almost all real systems where reinforcement learning is applied, it is found that a
knowledge free approach doesn't work. The basic RL algorithms must sufficiently be biased …

Reinforcement learning deals with learning optimal or near optimal policies while interacting
with an external environment. The applicability of reinforcement learning has been limited by …

Suppose we want to use an intelligent agent (computer program or robot) for performing
tasks for us, but we cannot or do not want to specify the precise task-operations. Eg we may …

In artificial intelligence the aim is to build intelligent entities (Russell and Norvig, 2009).
According to some, an artificial entity can be called intelligent when it can successfully mimic …

We present a method for automatically constructing macro-actions from scratch from
primitive actions during the reinforcement learning process. The overall idea is to reinforce …