Decision-Making and Learning in an Unknown Environment

Uwe Lorenz

Chapter 4 in Reinforcement Learning From Scratch, 2022, pp 47-121 from Springer

Abstract: Abstract This chapter describes how the agent can explore an unknown environmental system in which it has been placed. In doing so, he discovers states with rewards and has to optimize the paths to these goals, on the one hand, but also explore new goals, on the other hand. In doing so, he must consider a trade-off between exploitation and exploration. On the one hand, he has to collect the possible reward of already discovered goals; on the other, hand he has to manage the exploration of better paths or the discovery of new goals. There are different approaches to this; some aim at processing experiences made in such a way that the agent behaves better under the same conditions in the future “model-free methods”; and others that aim at optimizing models that can predict what would happen if certain actions are chosen.

Date: 2022
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:sprchp:978-3-031-09030-1_4

Ordering information: This item can be ordered from
http://www.springer.com/9783031090301

DOI: 10.1007/978-3-031-09030-1_4

Access Statistics for this chapter

More chapters in Springer Books from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().