Adaptive Representations for Reinforcement Learning - Shimon Whiteson
-30% with code BOOKS
Shipping in 12-18 days
30-day return policy
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 approach ... Full description
You May Also Like
TOP
If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All
Eliezer Yudkowsky, Nate Soares
-30% with code BOOKS
In stock at our supplier
TOP
How To Think About AI: A Guide For The Perplexed
-30% with code BOOKS
In stock at our supplier
TOP
Gödel, Escher, Bach: An Eternal Golden Braid
-30% with code BOOKS
In stock at our supplier
TOP
AI Engineering: Building Applications with Foundation Models
-30% with code BOOKS
In stock at our supplier
TOP
Empire of AI: Dreams and Nightmares in Sam Altman's OpenAI
-30% with code BOOKS
In stock at our supplier
TOP
Python Crash Course: A Hands-On, Project-Based Introduction to Programming
-30% with code BOOKS
In stock at our supplier
TOP
The Rust Programming Language
Steve Klabnik, Carol Nichols, Chris Krycho
-30% with code BOOKS
In stock at our supplier
TOP
How to Talk to AI: (And How Not To)
-30% with code BOOKS
In stock at our supplier
TOP
The Pragmatic Programmer: journey to mastery, 20th Anniversary Edition, 2/e: your journey to mastery, 20th Anniversary Edition
-30% with code BOOKS
In stock at our supplier
TOP
Hackers. 25th Anniversary Edition: Heroes of the Computer Revolution
-30% with code BOOKS
In stock at our supplier
TOP
The Art of Game Design: A Book of Lenses
-30% with code BOOKS
In stock at our supplier
TOP
How Linux Works: What Every Superuser Should Know
-30% with code BOOKS
In stock at our supplier
TOP
Hands-On Large Language Models: Language Understanding and Generation
Maarten Grootendorst, Jay Alammar
-30% with code BOOKS
In stock at our supplier
TOP
Fundamentals of Software Architecture: A Modern Engineering Approach
-30% with code BOOKS
In stock at our supplier
TOP
Speak Data: Artists, Scientists, Thinkers, and Dreamers on How We Live Our Lives in Numbers
-30% with code BOOKS
In stock at our supplier
TOP
Clean Architecture: A Craftsman's Guide to Software Structure and Design
-30% with code BOOKS
In stock at our supplier
TOP
World of Warcraft Chronicle, Volume 2
-30% with code BOOKS
In stock at our supplier
TOP
Linux Basics for Hackers, 2nd Edition: Getting Started with Networking, Scripting, and Security in Kali
-30% with code BOOKS
In stock at our supplier
TOP
Deep Learning: Foundations and Concepts
Christopher M. Bishop, Hugh Bishop
-30% with code BOOKS
In stock at our supplier
Description
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.
More Information
| Author | Shimon Whiteson |
|---|---|
| Publisher | Springer Berlin Heidelberg |
| Series | Studies in Computational Intelligence |
| Release year | 2014 |
| Cover type | Softcover |
| EAN | 9783642422317 |