Date: Tue, 26 Nov 1996 18:41:25 GMT
Server: NCSA/1.5.1
Last-modified: Tue, 26 Nov 1996 18:36:13 GMT
Content-type: text/html
Content-length: 16806
CS 152: Neural Networks
URL http://www.cs.hmc.edu/~keller/cs152f96.html
Trailer:
Can a computer be taught to read words aloud,
recognize faces, perform a medical diagnosis,
drive a car, play a game, balance a pole, predict physical phenomena?
The answer to all these is yes. All these applications and others
have been demonstrated using
varieties of the computational model known as "neural networks", the subject of this
course.
The course will develop the theory of a number of neural network models.
Participants will exercise the theory through both pre-developed computer
programs and ones of their own design.
Course Personnel:
- Instructor: Robert Keller
242 Olin (4-5 p.m. MTuW or by appt.), keller@muddcs, x 18483
- Tutor/Grader: T.J. Kelly
tkelly@muddcs, x 74860
- Secretary: Nancy Mandala 240 Olin (1-5 M-F) nancy@muddcs, x 18225
- System administrator: Quay Ly 101 Beckman quay@muddcs, x 73474
Catalog Description
Modeling, simulation, and analysis of artificial neural networks.
Relationship to biological neural networks. Design and optimization of
discrete and continuous neural networks. Backpropagation, and other gradient
descent methods. Hopfield and Boltzmann networks. Unsupervised learning.
Self-organizing feature maps. Applications chosen from function approximation,
signal processing, control, computer graphics, pattern recognition, time-series
analysis. Relationship to fuzzy logic, genetic algorithms, and artificial
life.
Prerequisites: Biology 52 and Mathematics 73 and 82, or permission of
the instructor. 3 credit hours.
Texts
- Main Textbook:
Martin T. Hagan,
Howard B. Demuth,
and Mark Beale,
Neural Network Design,
PWS Publishing Company, Boston, 1996, ISBN 0-534-94332-2,
which I will call NND below.
- Supplementary references, which will be provided as necessary.
- Related WWW links (for self-study and research)
appear below.
- Software, which is installed on muddcs.cs.hmc.edu:
Course Requirements
There will be some homework and programming
assignments, but no exams. These assignments will constitute about
50% of the grade. The other 50% of the grade is from a substantial
final project involving either a working neural network application or
a research paper. The grade on the project will be determined by the
comprehensiveness and degree to which you explored competing
approaches. The projects will be presented orally.
Optional
voluntary oral presentations on textbook material
can also be made during the term. These
can act to cushion your grade. They are very much encouraged, as it
they really help you learn the material at a higher level than you
would otherwise.
Please see me if you are interested in
making a presentation.
CS 152 Topic Outline
- Week 1 (read NND chapters 1 to 4; you may skip 3-8 to 3-12 for now)
Contexts for Neural Networks
- Artificial Intelligence
- Biological
- Physics
Artificial Neural Network overview
- Perceptrons
- Perceptron learning rule
- Perceptron convergence theorem
- Week 2 (read NND chapter 5 to 7)
- Linear transformations for neural networks
- Supervised Hebbian learning
- Pseudoinverse rule
- Filtered learning rule
- Delta rule
- Unsupervised Hebbian learning
- Week 3 (read NND chapters 8 and 9)
- Performance surfaces
- Performance optimization
- Steepest descent algorithm
- Newton's method
- Conjugate gradient
- Week 4 (read NND chapter 10)
Widrow-Hoff Learning
- Adaline
- LMS rule
- Adaptive filtering
- Week 5 (read NND chapters 11 and 12)
- Backpropagation in Multi-Level Perceptrons (MLP)
- Variations on backpropagation
- Batching
- Momentum
- Variable learning rate
- Levenberg-Marquardt (LMBP)
- Quickprop
- Week 6 (supplementary material)
- Radial basis function networks (RBF)
- Week 7 (read NND chapter 13)
- Associative learning
- Unsupervised Hebb rule
- Hebb rule with decay
- Instar rule
- Kohonen rule
- Outstar rule
- Week 8 (read NND chapter 14)
- Competitive networks
- Hamming network
- Self-Organizing feature maps (SOM)
- Counterpropagation networks (CPN)
- Learning vector quanitization (LVQ)
- Week 9 (read NND chapters 15 and 16)
- Grossberg networks
- Adaptive resonance theory
- Week 10 (read NND chapters 17 and 18)
- Hopfield networks
- Spin-glass model and simulated annealing
- Boltzman networks
- Cascade correlation learning
- Bi-directional associative memory (BAM)
- Week 11 (supplementary material)
- Sequential networks
- Time series
- Backpropagation through time
- Finite Impulse Response (FIR) MLP
- Method of temporal differences
The following additional, but related topics are assuming the previous
material hasn't expanded in time significantly. Whether it does
remains to be seen. If so, some of these topics may be compressed or
eliminated.
- Week 12
- Genetic programming and connection to NNs
- Other topics related to Alife (Artificial Life)
- Week 13
- Fuzzy logic and its connection to NNs
- Simon Haykin,
Neural networks - A comprehensive foundation,
Macmillan, 1994.
This book was used in the previous offering of the course.
It includes topics such as radial basis function networks and temporal
approaches which are not present in the main textbook. However the
mathematics is more difficult to follow.
- Mohamad H. Hassoun,
Fundamentals of artificial neural networks,
MIT Press, 1995.
This is another fairly thorough introduction.
- James A. Anderson,
An introduction to neural networks,
MIT Press, 1995.
This is a more gentle introduction to the topic, by one of the pioneers in the field.
-
Irwin B. Levitan and Leonard K. Kaczmarek,
The Neuron,
Oxford University Press, 1991.
This book focuses on the biology and physics of neurons.
-
Marvin L. Minsky and Seymour Papert,
Perceptrons (expanded addition),
MIT Press, 1988.
The historical importance of this book will be discussed in the course.
-
Duda and Hart,
Pattern classification and scene analysis,
Wiley, 1972.
This book gives a broad look at pattern classification problems, but is not
on neural nets as such.
-
Teuvo Kohonen
Self-organizing maps,
Springer-Verlag, 1995.
This is a comprehensive reference by the originator of this concept.
-
Bart Kosko,
Neural networks and fuzzy systems : a dynamical systems approach to machine intelligence,
Prentice Hall, 1992.
This book compares fuzzy and neural approaches to control problems.
-
Zbigniew Michalewicz,
Genetic Algorithms + Data Structures = Evolution Programs,
Third Edition,
Springer Verlag, 1996.
This book describes the evolutionary approach, which in some cases can achieve
results similar to neural approaches.
-
John R. Koza,
Genetic Programming,
MIT Press, 1994.
This book focuses on the evolutionary approach to producing programs.
-
Christopher G. Langton (ed.),
Artificial Life,
Addison-Wesley, 1989.
This is a colllection of articles on the topic.
Neurocomputing People
Software
Data
Demos