Output details
11 - Computer Science and Informatics
Middlesex University
Output title
Bounds of optimal learning
Type
E - Conference contribution
Name of conference/published proceedings
IEEE Symposium on Adaptive Dynamic Programming and Reinforcement Learning, 2009. ADPRL '09.
Volume number
-
Issue number
-
First page of article
199
ISSN of proceedings
-
Year of publication
2009
Number of additional authors
0
Additional information
<13> This paper gave rigorous answers to the following questions: 1) What is the minimum amount of information required to achieve a certain increase in performance? 2) What is the maximum increase in performance that is possible to achieve given a certain amount of information? These bounds are shown to be defined by the differences of certain information-theoretic potentials. The result was presented at the `IEEE International Symposium on Adaptive Dynamic Programming and Reinforcement Learning'. The work lead to the EPSRC project EP/H031936/1, where the theory was used to derive optimal mutation rates for evolutionary systems.
Interdisciplinary
-
Cross-referral requested
-
Research group
None
Citation count
0
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
