This book presents sequential decision theory from a
novel algorithmic information theory perspective. While the former
theory is suited for active agents in known environments, the
latter is suited for passive prediction of unknown environments.
The book introduces these two well-known but very different ideas
and removes the limitations by unifying them to one parameter-free
theory of an optimal reinforcement learning agent interacting with
an arbitrary unknown world. Most if not all AI problems can easily
be formulated within this theory, which reduces the conceptual
problems to pure computational ones. Considered problem classes
include sequence prediction, strategic games, function
minimization, reinforcement and supervised learning. Formal
definitions of intelligence order relations, the horizon problem
and relations to other approaches to AI are discussed. One
intention of this book is to excite a broader AI audience about
abstract algorithmic information theory concepts, and conversely
to inform theorists about exciting applications to AI.
%0 Book
%1 Hutter:04uaibook
%A Hutter, Marcus
%C Berlin
%D 2004
%I Springer
%K decision learning; ml prediction; reinforcement sequence sequential theory; universal
%P 300 pages
%T Universal Artificial Intelligence: Sequential Decisions based on Algorithmic Probability
%U http://www.hutter1.net/ai/uaibook.htm
%X This book presents sequential decision theory from a
novel algorithmic information theory perspective. While the former
theory is suited for active agents in known environments, the
latter is suited for passive prediction of unknown environments.
The book introduces these two well-known but very different ideas
and removes the limitations by unifying them to one parameter-free
theory of an optimal reinforcement learning agent interacting with
an arbitrary unknown world. Most if not all AI problems can easily
be formulated within this theory, which reduces the conceptual
problems to pure computational ones. Considered problem classes
include sequence prediction, strategic games, function
minimization, reinforcement and supervised learning. Formal
definitions of intelligence order relations, the horizon problem
and relations to other approaches to AI are discussed. One
intention of this book is to excite a broader AI audience about
abstract algorithmic information theory concepts, and conversely
to inform theorists about exciting applications to AI.
@book{Hutter:04uaibook,
abstract = {This book presents sequential decision theory from a
novel algorithmic information theory perspective. While the former
theory is suited for active agents in known environments, the
latter is suited for passive prediction of unknown environments.
The book introduces these two well-known but very different ideas
and removes the limitations by unifying them to one parameter-free
theory of an optimal reinforcement learning agent interacting with
an arbitrary unknown world. Most if not all AI problems can easily
be formulated within this theory, which reduces the conceptual
problems to pure computational ones. Considered problem classes
include sequence prediction, strategic games, function
minimization, reinforcement and supervised learning. Formal
definitions of intelligence order relations, the horizon problem
and relations to other approaches to AI are discussed. One
intention of this book is to excite a broader AI audience about
abstract algorithmic information theory concepts, and conversely
to inform theorists about exciting applications to AI.},
added-at = {2016-11-26T13:19:29.000+0100},
address = {Berlin},
author = {Hutter, Marcus},
bdsk-url-1 = {http://www.hutter1.net/ai/uaibook.htm},
biburl = {https://www.bibsonomy.org/bibtex/2191b593b677a3b539f0a734dafcfd7b9/machinelearning},
date-added = {2008-11-19 20:40:43 -0800},
date-modified = {2008-11-19 20:43:22 -0800},
interhash = {110c284b2e8ba1a6147e9cad94c1858b},
intrahash = {191b593b677a3b539f0a734dafcfd7b9},
keywords = {decision learning; ml prediction; reinforcement sequence sequential theory; universal},
pages = {300 pages},
publisher = {Springer},
timestamp = {2016-11-26T13:20:49.000+0100},
title = {Universal Artificial Intelligence: Sequential Decisions based on Algorithmic Probability},
url = {http://www.hutter1.net/ai/uaibook.htm},
year = 2004
}