WORCESTER POLYTECHNIC INSTITUTE
Computer Science Department

CS4341 ❏ Artificial Intelligence ❏ A'06

Mon, Tue, Thu, Fri - noon - AK 233
Prof. David C. Brown, Fuller Lab 131, (508) 831-5618, dcb at cs.wpi.edu

Version: Mon Sep 25 20:48:23 EDT 2006

PROJECT 3 - A Genetic Algorithm Learning System - Evaluation Criteria

Revised Draft

All projects will be graded out of 100 points for convenience, and will be adjusted later to conform to the class grading scheme already provided to you. Grading will be subtractive. That is, you start with 100 points, and points will be deducted for problems found. This produces lower scores, is harder to grade, but is much fairer and more consistent.

The grading will be divided into consideration of:
* 50 pts  Required       (i.e., what the problem description asked for)
* 10 pts  Presentation   (i.e., style, layout, comments)
* 40 pts  Demonstration  (i.e., the output from and actions by the system 
                                -- layout, clarity, completeness, how well tested).


REQUIRED:	

   - uses GA technique including mutation and crossover
   - uses rank-space method (quality rank and diversity rank)
   - uses external rule representation given in problem 
       (or very similar to it)
   - starts with one rule set (all values set to 1)
   - starts each new population with 4 rule sets
   - does all possible crossovers between the current rules sets
   - applies a single mutation transformation to each member of 
        the current population
   - doesnt allow duplicate rule sets in a population
   - uses table to evaluate fitness

   - have provided items: 

      # Brief, clear documentation that describes the design of your
        GA system. What special algorithms or data structures were
        used?

      # A description of, and rationale for, your rule set quality measure.
      # A description of, and rationale for, your rule set diversity measure.
      # A description of, and rationale for, your stopping condition.

      # The code for the system (which must be well commented).

      # The clear, readable output from the test/demonstration runs. 
      # These should not be annotated 
      # Some output in "demonstration" format.
      # output of condensed rule set representation for each generation


PRESENTATION:

        Clear documentation and descriptions.
        Good coding standards.
        Clear system output.
	Readable tests and explanations. 


DEMONSTRATION: 

        All aspects working & correct.
	How well tested (look for variations due to randomness).
	Completeness of messages in "demonstration" output
           (i.e., can we understand what's happening?)