Spring 2011: CS429/529: Introduction to Machine Learning
In this Section:
Introduction to Machine Learning
Lecture: Tues/Thurs, 12:30-1:45 PM
Mitchell Hall 211
Instructor: Terran Lane
The syllabus is now available.
If you're taking this class, you should probably subscribe to the ml-class mail list.
- Apr 29
- For anybody who is willing to share your slides with the rest of the class, I'll link them from the presentation schedule, below. I prefer if you just send me a URL to your slides (or bring them on a thumb drive or something) rather than emailing them, but I'll accept them by email if that's the only thing you can do.
- Apr 21
- The schedue for speakers for final project presentations is:
Presentation order Tues, Apr 26 Thu, Apr 28 Tues, May 3 Thu, May 5 1 D. Wadsworth M. Flynn G. Reedy J. Luan 2 T. Flanagan Z. Zhang D. Mo A. Ghaffarkhah 3 N. Salim T. Appel C. Steinebach D. Jaw 4 M. Hall T. Tangchoopong M. Martinez 5 C. Gunnning T. Adamson
- Apr 21
- Advice on the final project oral presentation is now available.
- Apr 7
- A number of announcements today:
- Class is canceled on Tues, Apr 12 and Thurs, Apr 14, due to Terran's travel.
- Homework 3 is still due Tues, Apr 12. You can email digital solutions or drop paper copies off at the department office (ask them to time stamp it and drop it in my mailbox).
- In lieu of class, we have Reading 5:
Yang, J., Xu, Y., and Chen, C.S. "Hidden Markov model approach to skill learning and its application to telerobotics". IEEE Trans. on Robotics and Automation, 10(5) pp 621-631. 1994. doi: 10.1109/70.326567
It should be available from on UNM campus.
Due: Tues, April 19.
NOTE:We assigned new reading groups in class today. If you missed class and need a new group assignment, please contact me.
- For Homework 3:
- You do not need to do any problems related to HMM parameter estimation/the Baum-Welch algorithm. (I.e., you do not need to do Part 3 of Question 4 or the parameter estimation part of Question 5.)
- Problem 5 is deferred to Tues, April 19. Christian has graciously volunteered to use off-the-shelf software to estimate a set of [A,B,pi] matrices for Q5. I will distribute those and you can use them to answer the two parts of this question.
- The HMM tutorial I mentioned in class is:
Rabiner, L., "A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition". Proc. IEEE 77(2), 1989. pp 257-286. doi: 10.1109/5.18626
This should be available through UNM via the libraries' full text article linker.
This article is essentially the same material as in the Rabiner & Juang textbook that I have been using for lecture. Its notation is close to that I used in class, except that I used Y for the hidden state, where Rabiner uses q, and I used X for the observed state, where Rabiner uses O.
- Mar 29
- Homework 3 is now available. Due: April 12.
- Mar 22
- The final project description is now available.
- Mar 8
- Reading 4:
Support Vector Machine Active Learning with Applications to Text Classification. Tong, S. and Koller, D. (2001). Journal of Machine Learning Reseach 2:45-66.Due: Thurs, Mar 24 (Thursday after Spring Break).
- Mar 3
- Homework 2 is now available. This is an ungraded homework, to give practice for the midterm.
- Mar 3
- The midterm exam will be in class, Thursday March 10, 2011.
- Feb 17
- Reading 3:
Optimal predictions in everyday cognition. Griffiths, T. L. and Tenenbaum, J. B. (2006). Psychological Science 17(9), 767-773.Due: Thurs, Jan 24.
- Feb 3
- Reading 2:
Caruana, R., "Multitask Learning". Machine Learning 28(1), 1997. pp 41--75. doi: 10.1023/A:1007379606734.Due: Feb 15
- Jan 25
- Homework 1 is now available.
- Jan 20
- The first reading is:
Samuel, A., "Some Experiments in Machine Learning Using the Game of Checkers". IBM J. of Research and Development, 3(3) pp 210-229. 1959. doi: 10.1147/rd.33.0210. This paper is not directly available through UNM libraries. We have access to two different scanned copies, though (for use only by members of the UNM CS429/529, Spring 2011 course):
- Jan 20
- Reminder: Reading 1 will be discussed in class on Thurs, Jan 27.
Each group should turn in a (less than or equal
to) one page analysis including:
- The names of all group members.
- One paragraph abstract of the paper. (Your own, novel abstract, not the paper's abstract.)
- At least one novel bit of analysis, including any subset of: ideas for refinements or extenstions, observations of mistakes or incomplete bits in the paper, open questions, suggestions for other applications for the material in the paper, questions about confusing bits, etc.