Fall 2008/2009 
CSC 578 Section 701 (Loop), 710 (DL)
Neural Networks and Machine Learning

Course Home Page: http://www.depaul.edu/~ntomuro/courses/578/

1. Course Description

A study of the basic structure of neural networks and how machines may learn. This will include analyses of decision trees, Bayesian learning, genetic algorithms, PAC, analytical and reinforcement learning. Neural networks to be studied include Backpropagation, Kohonen and Hopfield. Students will explore current applications and design several learning systems. No prior background in artificial intelligence is assumed. 

2. Pre-requisites

MAT 220 or MAT 262 or MAT 151.  This means a course in Linear Algebra or Calc II.  Students are also expected to be proficient in at least one programming language (Java, C++, C, Lisp, SAS, MatLab etc.)

3. Course Materials

Textbook (required):

• Machine Learning, Tom Mitchell, McGraw Hill, 1997, ISBN: 0070428077.
  

References (recommended; but not required):

• Data Mining: Practical Machine Learning Tools and Techniques (Second Edition), Ian H. Witten and Eibe Frank, Morgan Kaufman Publishers, 2005, ISBN: 0-12-088407-0.
  

4. Coursework

1. Assignments. There will be 4 assignments.  They are a mixture of written questions and experimentations using some publicly available machine learning tools.
2. Projects.  There will be 2 projects -- Midterm project and Final project.  Midterm project will be a computer implementation of a machine learning algorithm.  Students are free to program in any language.  Final project will be a research paper or a project.

5. Grading

The grade breakdown will be as follows:

    Assignments      55%
    Midterm Project  25%
    Final Project    20% 

The grading scale is as follows: 

    90-100       A
    80-89.99     B
    70-79.99     C
    60-69.99     D
     0-59.99     F

Plusses and minuses will be given at the high/low ends of each grade range (note: no A+'s).

• Grading Policy

• Your answers or programs will NOT be judged only by the final results/conclusions. Scores you receive will be based on the correctness of ALL parts of the answers or internal routines of the programs which are not displayed to the output.  So in some cases, points are taken off if an answer contains an incorrect statement, even though its final conclusion or output is correct.
• Assignments with NO NAME may be penalized by some points.

• Late Policy

  Assignments are due at 11:59 pm of the due dates.  Late assignments are accepted (through email attachment) with the following penalty. 

  If assignment is turned in...	Penalty will be...
  within 3 days of due date	10% of the total points for each day it is late
  3 days or more after due date	will NOT be accepted

• Plagiarism/Cheating and Incompletes

See the University and School policies on plagiarism and incompletes.

6. Tentative Schedule

The schedule will be updated throughout the quarter. Check the listing for readings and the dates of exams.

• Schedule