./02Logistics (revised)

 

This handout, given out January 27, is a revision of Handout 2.

 

1.    People

1.1.        Instructor

­          Dr. Andreas S. Weigend
Associate Professor of Information Systems

­          Office hours: after class, Tuesdays 2:00 - 3:00, and by appointment.
Location: K-MEC 9-74 (Information Systems)
Web: http://www.stern.nyu.edu/~aweigend
E-mail: aweigend@stern.nyu.edu
E-mail policy: I read text e-mail sent from a Stern account once a day. Short questions that require no more than a couple of lines of reply will be answered immediately. E-mail sent from non-Stern addresses and attachments are processed once a week.
Phone: 212 998-0803 (no voicemail, please use e-mail for messaging purposes)
Bio: PhD Stanford, 1991, joined Stern 1997, more in http://www.stern.nyu.edu/~aweigend/WeigendBio.html

­          If you have questions or feedback about the course, please contact me.

1.2.        Teaching assistant (TA)

­          Dr. Chenggang Shi

­          Lab hours: Mondays and Wednesday after class in UC70
E-mail: cs473@stern.nyu.edu
E-mail policy: Chenggang reads e-mail at least once a day.

­          If you have questions about the homework, please contact the TA.

1.3.        Administrative assistant

­          Hae-Sun Shon

­          Hours: 9:00am - 5:00pm.
Location: K-MEC 9-170 (Information Systems)
Phone: 212 998-0801
E-mail: hshon@stern.nyu.edu
Fax: 212 995-4228

­          If you missed a class and need a handout, or would like to borrow a videotape of a class, please contact the administrative assistant.

 

 

2.    Reading

2.1.        Books

2.1.1.     Learning from Data: Concepts, Theory, and Methods

­          Vladimir S. Cherkassky and Filip M. Mulier
Hardcover: USD75 (442 pages)
John Wiley & Sons (1998)
ISBN: 0471154938

­          Perspective: Statistics

2.1.2.     Predictive Data Mining

­          Sholom M. Weiss and Nitin Indurkhya
Paperback: USD40 (225 pages)
Morgan Kaufman Publishers (1997)
ISBN: 1558604030

­          Perspective: Business and Computer Science (Primarily verbal descriptions of problems and methods, includes preprocessing etc.)

2.1.3.     Neural Networks for Pattern Recognition

­          Christopher M. Bishop
Paperback: USD55 (482 pages)
Oxford University Press (1995)
ISBN: 0198538642

­          Perspective: Neural networks

2.2.        Additional readings

­          Some chapters from textbooks and some research papers are useful for this course. When available in electronic form, they will be placed in ./Readings/ in portable document format (.pdf) and/or postscript (.ps).  If I could not obtain an electronic copy from the authors or publishers, hardcopies will be distributed in class.

 

 

3.    Computers

3.1.        Course web site

­          The web site for this course is http://www.stern.nyu.edu/~aweigend/Teaching/DMF/S99 All links are given relative to this url and denoted by "./" .
There the 14-page proposal for this course that was awarded a 1988 NYU Curricular Development Challenge Fund is available as ./Oct98Original.html

­                                  The constantly updated schedule for this semester is in ./index.html

­          The homework assignments are in ./Homeworks/

­          The lecture notes are in ./Notes/

­          The reading material is in ./Readings/

­          Additional software is in ./Software/

3.2.        Course software

­          Most of the assignments for this course are done in MATLAB. For a first overview of the software, go to the website http://www.mathworks.com/products/matlab/

­          The software (version 5.2) has been installed on all computers in the UC70 lab. You can launch it from the Start Menu; it is under Information Systems.

­          For your convenience, the Mathworks also generously donated licenses and CD ROMs for all students in this course, allowing you to install the full software on your own computer. If you are interested in this option, please pick up a CD ROM from the administrative assistant. The pass code for your installation is 11113-04883-04484-19299-22792. The license expires on July 1, 1999. As minimum install, I recommend MATLAB, the Neural Network Toolbox, and the corresponding help files.

3.3.        Other useful software

­          If you know of more recent versions of any of these utilities, or better ones for similar purposes, please let me know and I will

3.3.1.     To edit text files

·         GNU Emacs

­          Emacs is a great text editor. Version 20.3.1 (August 1998) is described and can be downloaded from
http://www.cs.washington.edu/homes/voelker/ntemacs.html

·         Microemacs

­          Microemacs’98 (JASSPA distribution) (me32install.zip, 4MB) was downloaded from http://www.geocities.com/ResearchTriangle/Thinktank/7109/Evaluation

­          (An alternative, Microemacs 4.0 was released in 1996 and is harder to install.)

 

 

 

3.3.2.     To view and print postscript

­          Some of the readings are provided on the Web as postscript files.

·         ghostview

­          To preview and print postscript files.

­          Version 5.5 (gsv26550.exe, 3.3MB) was downloaded from
http://www.ate.uni-duisburg.de/Ghostscript/e.ghostscript.html

·         printfile

­          To print (but not preview) postscript and text files.

­          Version 2.1 (prfile21.zip, 0.2MB) was downloaded from
http://hem1.passagen.se/ptlerup/prfile.html

 

 

4.    Learning and Evaluation

4.1.        Learning

­          To get the most out of this class, the most important is that you review before each class what was done in the previous class. This is usually best done in a small group. I strongly encourage students to meet (for example just before or after class) on a regular basis with some other students in the class. I also welcome small groups of students coming as a team with their questions to the office hours.

4.2.        Evaluation

4.2.1.     Homeworks

­          There will be regular homeworks that need to be handed in and will be graded by the TA. The number of points given for each homework reflects the amount of work its solution typically requires.

4.2.2.     Quizzes and self-tests

­          In the interest of not losing class time, I am not giving quizzes in this course. I will, however, hand out some self tests on important concepts from the previous class. You should have no problem to write down the answers to those.

4.2.3.     Final

­          The final is based on the material covered in class, on the homeworks and on the self tests. The date is given in the section on dates and times.

4.2.4.     Project

­          I am not requiring individual projects, but they are possible for extra credit. More importantly, working on a real problem you are genuinely interested in with real data is an extremely important learning experience. If you have suggestions for a project and for the corresponding data, please contact me about it. I prefer projects to be done in small groups, and it is often a good learning experience to have members with heterogeneous backgrounds in your group.

 

5.    Dates and Times

5.1.        Classes and final

­          Class time: Mondays and Wednesdays, 7:00 - 8:20pm. Please be on time.
Classroom: K-MEC 2-70.

­          No class on Monday, February 15 (Presidents' Day)

­          No class on Monday, March 15 and Wednesday, March 17 (Spring Break)

­          Full day workshop: Saturday, April 10 (begins 9:00am with breakfast)

­          Final exam: Monday, May 10, 7:35 - 9:35pm.

5.2.        University deadlines

­          Last day to add: January 28

­          Last day to drop (without "W"): February 16

­          Last day to withdraw from classes (with a "W"): March 29

6.    Additional course (after Spring Break)

­          I am teaching a small seminar (B20.3389) after Spring break. While it primarily targets PhD students in Information Systems, I welcome students from DMF who are interested in learning more and understanding the issues more deeply by discussing some of the most exciting and promising recent research. This course will covers the following topics:

o        1. Neural Networks (week of Mar 22)
Paper by Simard et al

o        2. and 3. Reinforcement learning (weeks of Mar 29 and Apr 5)
Tutorial by van Roy
Paper by Moody et al
Paper by Neuneier et al

o        4. Association rules (week of Apr 12)
Paper by Agrawal
Paper by Tuzhilin et al

o        5. Evolutionary computing (week of Apr 19)
Genetic algorithms paper by Dhar et al
Genetic programming paper by Chidambaran et al

o        6. and 7. Bayes nets, graphical models, causal networks, influence diagrams,
Tuturial by Buntine
Tutorial by Heckerman
Possibility of a project using the Hugen software

­          Please talk to me in person if you are interested in participating in this special course.