CPS 271: Numeric Artificial Intelligence
Fall 2003
Course Info
- Although titled "Numeric Artificial Intelligence", this course is an introduction to machine learning. It is a qual course.
- Meeting Times: Tuesday & Thursday, 12:40 - 1:55, LSRC
D106
- Instructor: Dr. Ron Parr - parr@cs.duke.edu,
D209 LSRC, 660-6537
Office Hours: Tuesday/Thursday, 1:55 - 2:55
- TA: Paul Shealy - pauls@cs.duke.edu,
D224 LSRC, 660-6583
Office Hours: Tuesday 11:00 - 12:00, Thursday 3:00 - 4:00
- Textbook: Neural Networks for Pattern Recognition, Christopher
M. Bishop, Oxford Press
- Alternate Texts: Introduction to Computational
Learning Theory, Michael Kearns & Umesh Vazirani, MIT Press
Machine Learning, Tom Mitchell, McGraw-Hill
The Elements of Statistical Learning, Trevor Hastie, Robert
Tibshirani, and Jerome Friedman, Springer Verlag
- Final Exam: The final exam is Friday, December 12, 9:00 AM - 12:00 PM, per the Registrar's exam schedule.
- Midterm: The midterm will be October 16th.
Grading
- homework: 10%
- midterm: 30%
- project: 30%
- final: 30%
A qual pass/fail is based only on the midterm and final.
Tentative syllabus
- Review of Probability
- Review of Statistics and Linear Algebra
- Statistical Pattern Recognition
- Unsupervised Learning
- Probability Density Estimation
- Hidden Markov Models
- Basic Neural Networks and Decision Trees
- Reinforcement Learning
- Bias and Variance
- Bayesian Techniques
- Computational Learning Theory
- Support Vector Machines
Assignments
Homework:
- Homework #1 - ps, pdf - due Thursday, 9/11/03
- Homework #2 - ps, pdf
- due Monday, 9/29, 5PM
- Homework #3 - ps, pdf - due Wednesday, 9/15, 5PM
- Homework #4 - ps, pdf - due Friday, 11/14
- Homework #5 - ps, pdf - due Tuesday, 12/2
Requirements for the final project handin, due December 5th: ps, pdf.
Other
Links:
- The k-means demo shown in class can be found here.
- The perceptron demo can be found here.
- Kevin Murphy's HMM page can be found here.
- the Sutton & Barto text "Reinforcement Learning: An Introduction" can be found here.
- The page on eigenfaces can be found here .
- A tutorial on PCA in Matlab.
- Two excellent tutorials on support vector machines are those by Hearst (ed.) and Burges.
- An SVM demo can be found here.
- A brief overview paper
on boosting and a more detailed paper.
- Pages on boosting and SVMs.
Slides: