Fetching latest commit…
Cannot retrieve the latest commit at this time.
| Failed to load latest commit information. | |||
|
conf | ||
|
|
doc | ||
|
|
graph | ||
|
|
log | ||
|
|
src | ||
|
|
tutorial | ||
|
|
.gitignore | ||
|
|
Makefile | ||
|
|
graph.py | ||
|
|
license.txt | ||
|
|
mc-aixi-ctw.sln | ||
|
|
mc-aixi-ctw.suo | ||
|
|
mc-aixi-ctw.vcxproj | ||
|
|
mc-aixi-ctw.vcxproj.filters | ||
|
|
mc-aixi-ctw.vcxproj.user | ||
|
|
readme.txt | ||
readme.txt
This software package consists of a simple implementation of MC-AIXI-CTW, an intelligent agent that learns from experience how to perform well in a wide variety of environments. This includes, but is not limited to the example games provided in this package, such as Tic Tac Toe, Pacman, and Kuhn Poker. The package consists of: * tutorial: contains a brief tutorial to using the MC-AIXI-CTW software * src: contains the source code for the agent * conf: contains configuration files * doc: contains the automatically generated documentation doc/html/index.html: starting point for HTML documentation * graph: contains generated graphs * log: contains log files. The project comes with log files generated by allowing the agent to explore for 5000 cycles then evaluating the performance of the learned policy for another 5000 cycles. The latest version of this software can be downloaded from http://www.hutter1.net/publ/aixictwxcode.zip A scientific motivation and description of MC-AIXI-CTW can be found in J.Veness, K.S.Ng, M.Hutter, W.Uther, D.Silver, A Monte Carlo AIXI Approximation, Journal of Artificial Intelligence Research, 40 (2011) 95-142 http://dx.doi.org/10.1613/jair.3125 Free TechReport version: http://arxiv.org/abs/0909.0801 BibTeX: http://www.hutter1.net/official/bib.htm#aixictwx