This page is designed to be a useful resource for students
taking the COGS 535 course (Computational Issues in Cognitive Science). Contributions
to this page are very welcome. Please mail them to
istvan@louisiana.edu
.
- The SOAR Home Page
. Versions of the SOAR architecture can be downloaded from here. Versions
are available for both Windows and Linux. Examples are included with the
software. There is a graphical version of the 8-puzzle, for instance.
- The Python Home Page
. The Python programming language can be downloaded from this page. In addition,
a variety of documents and guides are also available.
- CMU AI Repository
. This page provides links to the source code for all sorts of classic AI
systems and architectures. It is well worth exploring some of the systems
available from this resource.
- SWI-Prolog Home Page
. This dialect of prolog is available on UCS UNIX computing systems. To invoke
SWI-Prolog, use the command 'pl'. The alias 'prolog' will bring up a vary
old version of the language, that is not recommended. The SWI-Prolog page
provides lots of useful information and documentation.
- John McCarthy's Home Page
. John McCarthy is the 'father' of the language LISP and made many influential
contributions to the study of Artificial Intelligence. His home page has
many useful and interesting links on it, including links to papers by him.
- Paradigms of AI Programming Source Code
. This page has LISP source code for a number of classic Artificial Intelligence
programs, including a couple of version of ELIZA and the General Problem
Solver. If you follow the instructions on the page, it is possible to run
these programs on UCS UNIX systems. Be sure to exactly follow the instructions
on the page however. In order to invoke LISP whilst logged onto UCS UNIX
systems, just use the command 'lisp', then follow the instructions on the
page.
- AI on the Web
. This page has over 800 links to AI resources available on the web. The
page is organised into particular topics and each topic is then broken down
into further sub-categories. This page is a very useful resource where some
exploring can throw up lots of interesting information.
- Telerobot
. This is rather a cute and interesting page that enables you to control a robotic arm in Australia via a Java interface.
- The UC
Irvine Machine Learning Database web page has some resources that
can be useful. It contains some classic machine learning problem sets.
- The Artificial Neural Network Analysis Project Lab, Neural Network Software.
This suite of software was developed under a Board of Regents Research
Competitiveness Subprogram Grant. It will enable you to train and analyse
Artificial Neural Networks (ANNs).
- The ANNAP software manual can be found at http://www.ucs.louisiana.edu/~isb9112/dept/ANNAPCode/Docs/manual/
- The ANNAP software runs in the MATLAB
environment. MATLAB is available on UL Lafayette UCS computing systems.
- To see the source code for the software, log into your UCS account.
At the command prompt issue the command 'cd /w1/isbres/ANNAPCode'. The basic
ANN functions can be found in the Base/ subdirectory. Useful tools can be
found in the Tools/ subdirectory. The manual can be found in the Docs/manual/
subdirectory. Finally, some files for running examples can be found in the
examples/ subdirectory.
- To invoke the MATLAB environment from a UNIX login shell, just type the command 'matlab' at the command prompt on UCS syste
ms. Note, MATLAB can be slow to launch, so be patient. Also, you may get
an error message, type 'y' if you are asked to ignore this error. MATLAB
will load after a few seconds.
- The problem with using the shell approach to running MATLAB is that
the graphics based functions do not work. To make these functions work, you
will need an XWindow interface. A good XWindow interface can be downloaded
from http://www.xmanager.com . (N.B.
This will give you access to a thirty day trial version). You are strongly
encouraged to download and install this software on your computer. Once the
software is installed, run 'X Start', giving a UCS machine (e.g. s53.ucs.louisiana.edu)
as the machine you wish to connect to. In the 'Execute Command' window
, the command to run is 'xterm -ls -display $DISPLAY' (i.e. delete all the
path information before 'xterm'). You shoul.d then be able to make a graphical
connection to UCS systems and be able to take full advantage of MATLAB. To
run MATLAB, just type the command 'matlab' at the command prompt. Note, it
can take quite a while for the software to launch, especially if you are using
a modem connection, or if the workstation you are logged into is very busy,
so you may have to be patient!
- In order to use the ANNAP software, you will need to set up a directory for control files, and the input and output d
ata. First, make sure that you are in your home directory on UCS. You
can do this by issuing the command 'cd ~/'. Next you need to create a special
subdirectory for your ANN simulations. You can do this by issuing the command
'mkdir ANNSims'. There are two things to note here; (1) You can replace the
string 'ANNSims' with anything you like. (2) UNIX is case sensitive, so it
will treat a directory called 'ANNSims' as being different from a directory
called 'annsims'. It is assumed that you have choosen to call your directory
'ANNSims', if you have named it something else, then adjust commands appropriately.
- The next step is to copy the required files into your ANNSims directory.
To do this, issue the command 'cp /w1/isbres/ANNAPCode/examples/*.* ~/ANNSims/'.
-
Once you have copied the file, look at the file readme.txt for further instructions.
A simple way to do this is to issue the command 'more readme.txt'.