- FACULTY and STAFF
- RESEARCH AREAS
- APPLY NOW
- GIVE TODAY
Graduate Course Information
ECE 631 - Neural Networks
Course Website: D2L
UA Catalog Description: http://catalog.arizona.edu/allcats.html
Homework: 8 – 10 assignments
Project: 1 project
Exams: 2 Midterm Exams, 1 Final Exam
Typically: 40% Midterms,
40% Final Exam,
Course Summary: Neural networks represent a novel form of computing that is motivated by the highly parallel processing that takes place in biological systems. The course will present the theory and application of parallel distributed computation via elementary processing elements; neuron models and biological analogies; relationships between neural networks and statistical classification, supervised/unsupervised learning; neural net models; associative memories; training algorithms
Required: Simon Haykin, “Neural Networks: A Comprehensive Foundation,” 2nd Edition, Prentice Hall, 1999. Although this text will be required for the course, significant supplementary materials will be presented as class notes. The following list of references may also prove helpful for occasionally consultation.
Duda and Hart, “Pattern Analysis and Scene Classification,” John Wiley & Sons, 1973.
Bishop, “Neural Networks for Pattern Recognition,” Oxford University Press, 1995.
Cherkassky and Mulier, “Learning From Data,” John Wiley & Sons, 1998.
1. Introduction to Neurocomputing (CH1)
2. Learning and Generalization (CH2)
3. The Neuron (CH3)
4. Multi-Layer Networks (CH4)
5. Miscellaneous Network Studies
a. - Regularization Nets (CH5)
b. - Competitive Nets (CH9)
c. - Support Vector Machines (CH6)
6. Feedback and Optimization Nets (CH14)
Lecture: 150 minutes/week