Agent Demonstrator
==================

Written by Derek Bridge for CS4071 Artificial Intelligence.

1. Download and unpack the file. (Separate instructions on
   the course web page explain how to do this.)

2. Change directory:

   cd agentDemonstrator

3. The files are already compiled. If for some reason they
   need to be recompiled (e.g. if you have changed them),
   then you can compile each subdirectory in turn, as follows:

   javac actions/*.java
   javac demos/*.java

   etc.

4. Now you're ready to run the system:

   java demos.Main
   - this will run the system with a blank world to which
     you can add bricks, randomly moving lights, light
     seeking agents, etc.

   java demos.WallFollowerMain
   - this will run the system with a world that contains
     some walls and you can add a wall following agent

   java demos.NeuralWallFollowerMain
   - this will train a neural net to do wall following and
     then display a walled world and an agent whose
     behaviour is determined by the neural net

   java demos.GAWallFollowerMain 1000 7 0.8 0.1 0.1 15
   - this will evolve a production system for wall following
     and then display a walled world and an agent whose
     behaviour is determined by the evolved production system.
     (If you want to know wht the 6 numbers stand for, try to
     run the program without supplying the numbers. An error
     message will explain what the 6 numbers are.)


  
