Ohio University
School of Electrical Engineering and Computer Science
EE690- Design of self organizing neural networks
hardware and software issues
Course Schedule, Spring 2001

Updated March 27, 2001

Wk 
Date 
Subject 
Reading 
Homework
1 
Introduction
March 26 Class and lab organization
Questions about neural networks		 notes
Neural Networks
Discussion of the self-organizing NN architecture (SONNAR) Applets for Neural Networks and Artificial Life Matlab project
see also Project discussion
Threshold selection in a SONNAR KOHONEN'S Neural Networks
Kohonen nets		 Review reference books for presentation topic
2
 April 2 Approximation of the information index Bayesian Self-Organizing Maps
Bayesian Self-Organizing Map Simulation		 read ch. 1 pp.1-8
Matlab code selforgNN.m
NNorgVote.m dpyAIINNInfo.m
plot_prewired_structure.m
plot_learned_structure.m,  plotcut.m
scale_data.m  maint0.m
Fast learning process in a SONNAR NN toolbox read ch.2 pp.9-28
Discussion of the NN Matlab code Evolving hardware HW 1 due
read ch.2 pp. 29-40
3
 April 9 genesis1
genesis 2		 Matlab project training 0, 1, 2,   3,  and  4      data
 test on the training data 
Cellular neural networks concept notes Submit a topic for classroom presentation
Check the proposal requirements here
HW 2 due
4
 April 16 Basic circuits for VLSI neural networks VLSI NN circuits read ch.3 pp. 41-56
Neural and electrical systems
Gilbert transconductance multiplier		 Digital Self-organizing NN Matlab project competition - show test results on all data sets
save training results
Analog Self-organizing NN Matlab project competition - measure correct classification on test set
5
 April 23 notes, synapses HW 3 due
read ch.3 pp. 56-76
Matlab project competition - measure design complexity, training time and performance
6
 April 30 HW 4 due clustering data
NN companies read ch.4 pp. 76-96
7
 May 7 No classes this week Tactile Sensors HW 5 due
Evolvable NN
Molecular Computing
8
 May 14  NEuroNet  Matlab project competition -
functions must be normalized, different design parameters explored
9 
 May 21 HW 6 due write a report based on one of the following (each student chose different paper):
HARDWARE NEURAL NETWORKS
Digital NN
Microprocessor for NN
Neural Network Hardware
10 
 May 28 Classroom Presentations
June 2 Final Projects Due
Notes: Chapter sections are from the text by Marc M. Van Hulle, "Faithful Representations and Topographic Maps: From Distortion- to Information-Based Self-Organization" John Wiley & Sons; 2000.
and a reference by C. Mead, "Analog VLSI and Neural Systems", Addison-Wesley, 1989,

Homeworks are due by the date indicated.