Results 1  10
of
1,105,810
Learning MultiGoal Dialogue Strategies Using Reinforcement Learning With Reduced StateAction Spaces
 International Journal of Game Theory
, 2006
"... Learning dialogue strategies using the reinforcement learning framework is problematic due to its expensive computational cost. In this paper we propose an algorithm that reduces a stateaction space to one which includes only valid stateactions. We performed experiments on full and reduced spaces ..."
Abstract

Cited by 12 (2 self)
 Add to MetaCart
Learning dialogue strategies using the reinforcement learning framework is problematic due to its expensive computational cost. In this paper we propose an algorithm that reduces a stateaction space to one which includes only valid stateactions. We performed experiments on full and reduced spaces
Reinforcement Learning in Multidimensional StateAction Space using Random Rectangular Coarse Coding and Gibbs Sampling
"... Abstract — This paper presents a coarse coding technique and an action selection scheme for reinforcement learning (RL) in multidimensional and continuous stateaction spaces following conventional and sound RL manners. RL in highdimensional continuous domains includes two issues: One is a general ..."
Abstract
 Add to MetaCart
Abstract — This paper presents a coarse coding technique and an action selection scheme for reinforcement learning (RL) in multidimensional and continuous stateaction spaces following conventional and sound RL manners. RL in highdimensional continuous domains includes two issues: One is a
The irreducibility of the space of curves of given genus
 Publ. Math. IHES
, 1969
"... Fix an algebraically closed field k. Let Mg be the moduli space of curves of genus g over k. The main result of this note is that Mg is irreducible for every k. Of course, whether or not M s is irreducible depends only on the characteristic of k. When the characteristic s o, we can assume that k ~ ..."
Abstract

Cited by 512 (2 self)
 Add to MetaCart
Fix an algebraically closed field k. Let Mg be the moduli space of curves of genus g over k. The main result of this note is that Mg is irreducible for every k. Of course, whether or not M s is irreducible depends only on the characteristic of k. When the characteristic s o, we can assume that k
Antide Sitter Space, Thermal Phase Transition, and Confinement in Gauge Theories
 Adv. Theor. Math. Phys
, 1998
"... The correspondence between supergravity (and string theory) on AdS space and boundary conformal field theory relates the thermodynamics of N = 4 super YangMills theory in four dimensions to the thermodynamics of Schwarzschild black holes in Antide Sitter space. In this description, quantum phenome ..."
Abstract

Cited by 1087 (4 self)
 Add to MetaCart
The correspondence between supergravity (and string theory) on AdS space and boundary conformal field theory relates the thermodynamics of N = 4 super YangMills theory in four dimensions to the thermodynamics of Schwarzschild black holes in Antide Sitter space. In this description, quantum
StateAction Aggregations for Reinforcement Learning
, 2006
"... This paper offers an approach to the problem of large state spaces for reinforcement learning by constructing a stateaction pair aggregation (treating similar stateaction pairs as if they were the same) with the use of domain knowledge. Arbitrary aggregation is known to give possibly very large er ..."
Abstract
 Add to MetaCart
This paper offers an approach to the problem of large state spaces for reinforcement learning by constructing a stateaction pair aggregation (treating similar stateaction pairs as if they were the same) with the use of domain knowledge. Arbitrary aggregation is known to give possibly very large
Learning realistic human actions from movies
 IN: CVPR.
, 2008
"... The aim of this paper is to address recognition of natural human actions in diverse and realistic video settings. This challenging but important subject has mostly been ignored in the past due to several problems one of which is the lack of realistic and annotated video datasets. Our first contribut ..."
Abstract

Cited by 714 (51 self)
 Add to MetaCart
next turn to the problem of action classification in video. We present a new method for video classification that builds upon and extends several recent ideas including local spacetime features, spacetime pyramids and multichannel nonlinear SVMs. The method is shown to improve state
Using Expert Knowledge to Construct Error Bound StateAction Aggregations for Reinforcement Learning
"... Reinforcement learning is a wellsuited approach for many decisionmaking problems. Lots of interesting domains are, however, not solvable in practice by this approach due to their size: traditional reinforcement learning algorithm need to store every combination of state and action which was encoun ..."
Abstract
 Add to MetaCart
was encountered. A common method for dealing with large state spaces consists of partitioning the state space. Arbitrary partitioning can, however, introduce very large errors. In this paper we introduce an approach to construct a partitioning of the stateaction space with a provable error bound, using expert
Fitted natural actorcritic: A new algorithm for continuous stateaction MDPs
, 2008
"... In this paper we address reinforcement learning problems with continuous stateaction spaces. We propose a new algorithm, tted natural actorcritic (FNAC), that extends the work in [1] to allow for general function approximation and data reuse. We combine the natural actorcritic architecture [1] ..."
Abstract

Cited by 4 (1 self)
 Add to MetaCart
In this paper we address reinforcement learning problems with continuous stateaction spaces. We propose a new algorithm, tted natural actorcritic (FNAC), that extends the work in [1] to allow for general function approximation and data reuse. We combine the natural actorcritic architecture [1
CMACs for Representing StateAction Functions
"... Cerebellar Model Articulation Controllers (CMACs, [1–4]) are used to represent the stateaction value function Q: S×A → by a function approximator ˆ Qw with realvalued parameter vector w. In our implementation, a CMAC neural network is based on c regular tilings. Each tiling partitions the input sp ..."
Abstract
 Add to MetaCart
Cerebellar Model Articulation Controllers (CMACs, [1–4]) are used to represent the stateaction value function Q: S×A → by a function approximator ˆ Qw with realvalued parameter vector w. In our implementation, a CMAC neural network is based on c regular tilings. Each tiling partitions the input
Results 1  10
of
1,105,810