This program implements Q-learning on a given MDP, using 
semi-uniform exploration.

Requirements:
	GNU C compiler

To install:

1. Set values: (the defaults may be used for a quick start)

   a.) In the file mdp-qlearner.c, set the reinforcement learning values:

     - States and actions - these numbers must correspond to the
       probability and reward matrices defined in the files mpd-prob and
       mdp-reward.  
     - Number of iterations and step size
     - Exploration threshold
     - Number of runs
     - Learning rate (BETA)
     - Discount factor (GAMMA)

   b.) In the file mdp-prob, type in the probability matrix.

   c.) In the file mdp-reward, type in the reward matrix.

2. Type gmake.  The executable will be called "mdp-q"

3. Run "mdp-q".  You will be shown (in text mode) the trials. 

   
