Server: Netscape-Commerce/1.12 Date: Wednesday, 20-Nov-96 23:23:34 GMT Last-modified: Sunday, 03-Mar-96 02:27:20 GMT Content-length: 3246 Content-type: text/html EE539 Home Page
EE539: Neural Networks & Applications
Nabil H. Farhat

Table of Contents

Brief Description
General Information
Course Syllabus
Textbook
Grading Policy
Exam Dates



Course Description

Examines the application of paradigms in neural networks to problems in pattern classification, optimization, function approximation, and machine learning. The course will include: review of the physiology and anatomy of neurons and neuron networks, form al models of neurons and networks; attractor networks, associative memory; storage capacity; the pattern classification problem; neural classifiers; optimization by energy minimization, solving the TSP (Traveling Salesman problem) with attractor networks; simulated annealing and the Boltzmann machine; hardware implementations of neural networks; the problem of learning; algorithmic approaches; perceptron learning; back-propagation; randomized algorithms; and genetic algorithms.

General Information

    Nabil H. Farhat

    Nabil H. Farhat
    Room 372 Moore
    Phone: 898-5882
    Email:farhat@ee.upenn.edu

    Office Hours

    If unavailable, please see, Drucilla Spanner, Room 363 Moore, 898-6823

    Prerequisite

    None (Undergraduates need permission of Instructor)

    Time and Location

    TTh, 3-4:30, 223 Moore

Course Syllabus

  • Topic 1, Review of Essential Properties of the Biological Neuron and the Nervous System
  • Topic 2, Essentials of Nonlinear Dynamical System Theory
  • Topic 3, The Hopfield Model and Spin Glasses
  • Topic 4, Stochastic Neural Networks and the Boltzmann Machine
  • Topic 5, Multilayer Feedforward Networks for Supervised Learning
  • Topic 6, Unsupervised and Competitive Learning Algorithms
  • Topic 7, Bifurcating Neural Networks

Textbooks

    Main texts

    1. Neural Network Architectures , Dayhoff, J., Van Nostrand Reinhold, 1990.

    Reference

    1. Neural Computing: Theory and Practice, Wasserman, Philip, D., Van Nostrand Reinhold, 1989
    2. Introduction to the Theory of Neural Computation Hertz, J., Krogh,A., and Palmer, R.G., Addison Wesley, 1991
    3. Supplementary classnotes and material for additional reading will be handed out in class.

Grading Policy

  • Homeworks: 1/3
  • Midterm: 1/3
  • Final: 1/3

Exam Dates

  • Midterm 1: TBA
  • Final: Mo.Dec. 18, 1:30-3:30 pm

Nabil H. Farhat
Updated: Sept. 21, 1995