Cargando…

Adaptive Representations for Reinforcement Learning

This book presents new algorithms for reinforcement learning, a form of machine learning in which an autonomous agent seeks a control policy for a sequential decision task. Since current methods typically rely on manually designed solution representations, agents that automatically adapt their own r...

Descripción completa

Detalles Bibliográficos
Clasificación:Libro Electrónico
Autor principal: Whiteson, Shimon (Autor)
Autor Corporativo: SpringerLink (Online service)
Formato: Electrónico eBook
Idioma:Inglés
Publicado: Berlin, Heidelberg : Springer Berlin Heidelberg : Imprint: Springer, 2010.
Edición:1st ed. 2010.
Colección:Studies in Computational Intelligence, 291
Temas:
Acceso en línea:Texto Completo

MARC

LEADER 00000nam a22000005i 4500
001 978-3-642-13932-1
003 DE-He213
005 20220117202342.0
007 cr nn 008mamaa
008 100709s2010 gw | s |||| 0|eng d
020 |a 9783642139321  |9 978-3-642-13932-1 
024 7 |a 10.1007/978-3-642-13932-1  |2 doi 
050 4 |a Q342 
072 7 |a UYQ  |2 bicssc 
072 7 |a TEC009000  |2 bisacsh 
072 7 |a UYQ  |2 thema 
082 0 4 |a 006.3  |2 23 
100 1 |a Whiteson, Shimon.  |e author.  |4 aut  |4 http://id.loc.gov/vocabulary/relators/aut 
245 1 0 |a Adaptive Representations for Reinforcement Learning  |h [electronic resource] /  |c by Shimon Whiteson. 
250 |a 1st ed. 2010. 
264 1 |a Berlin, Heidelberg :  |b Springer Berlin Heidelberg :  |b Imprint: Springer,  |c 2010. 
300 |a XIII, 116 p.  |b online resource. 
336 |a text  |b txt  |2 rdacontent 
337 |a computer  |b c  |2 rdamedia 
338 |a online resource  |b cr  |2 rdacarrier 
347 |a text file  |b PDF  |2 rda 
490 1 |a Studies in Computational Intelligence,  |x 1860-9503 ;  |v 291 
505 0 |a Part 1 Introduction -- Part 2 Reinforcement Learning -- Part 3 On-Line Evolutionary Computation -- Part 4 Evolutionary Function Approximation -- Part 5 Sample-Efficient Evolutionary Function Approximation -- Part 6 Automatic Feature Selection for Reinforcement Learning -- Part 7 Adaptive Tile Coding -- Part 8 RelatedWork -- Part 9 Conclusion -- Part 10 Statistical Significance. 
520 |a This book presents new algorithms for reinforcement learning, a form of machine learning in which an autonomous agent seeks a control policy for a sequential decision task. Since current methods typically rely on manually designed solution representations, agents that automatically adapt their own representations have the potential to dramatically improve performance. This book introduces two novel approaches for automatically discovering high-performing representations. The first approach synthesizes temporal difference methods, the traditional approach to reinforcement learning, with evolutionary methods, which can learn representations for a broad class of optimization problems. This synthesis is accomplished by customizing evolutionary methods to the on-line nature of reinforcement learning and using them to evolve representations for value function approximators. The second approach automatically learns representations based on piecewise-constant approximations of value functions. It begins with coarse representations and gradually refines them during learning, analyzing the current policy and value function to deduce the best refinements. This book also introduces a novel method for devising input representations. This method addresses the feature selection problem by extending an algorithm that evolves the topology and weights of neural networks such that it evolves their inputs too. In addition to introducing these new methods, this book presents extensive empirical results in multiple domains demonstrating that these techniques can substantially improve performance over methods with manual representations. 
650 0 |a Computational intelligence. 
650 0 |a Artificial intelligence. 
650 1 4 |a Computational Intelligence. 
650 2 4 |a Artificial Intelligence. 
710 2 |a SpringerLink (Online service) 
773 0 |t Springer Nature eBook 
776 0 8 |i Printed edition:  |z 9783642422317 
776 0 8 |i Printed edition:  |z 9783642139314 
776 0 8 |i Printed edition:  |z 9783642139338 
830 0 |a Studies in Computational Intelligence,  |x 1860-9503 ;  |v 291 
856 4 0 |u https://doi.uam.elogim.com/10.1007/978-3-642-13932-1  |z Texto Completo 
912 |a ZDB-2-ENG 
912 |a ZDB-2-SXE 
950 |a Engineering (SpringerNature-11647) 
950 |a Engineering (R0) (SpringerNature-43712)