Advanced Machine Learning (CSCI 6622)

 

Course offered spring term 2004.

 

Location:          

Monday and Wednesday 03:00pm-04:15pm ECCR 116

Instructor:

Professor Greg Grudic

Office:            

ECOT 525

Office Hours:

Tuesday and Wednesday 4:30-5:30

Phone:

303-492-4419

Email:

grudic@cs.colorado.edu

Course URL:

http://www.cs.colorado.edu/~grudic/teaching/CSCI6622_2004

 

 

How does an internet vender know which consumer group or individual to target for advertising? How do we find all pictures of Arnold Schwarzenegger in a large image database? Which gene is responsible for the cancer that runs in my family? Machine learning algorithms are at the heart of many computer applications designed to address these types of questions. Other important applications include Data Mining, Robotics, Profiling, User Interfaces, Document Characterization, Bioinformatics, and Linguistics.

 

Machine learning is the study of building systems that learn from experience. Machine learning algorithms are designed to address problem domains where good theoretical models don’t exist, but where empirical observations can be made.

 

This course is designed for two types of students:

 

  1. Students who are interested in applying state of the art machine learning algorithms to build models from data that cannot easily be understood theoretically. This includes both advanced Ph.D. students who have data they are trying to understand (e.g. linguistics, robotics, physics etc.), as well as Masters Students who have a problem domain they are particularly interested in.
  2. Students who are interested in formulating new machining learning algorithms, or analyzing the theory behind existing ones.

 

In bringing together students interested in both application and theory, I hope to create a rich research environment which will benefit both groups.

 

The goal of this course is to do original and interesting research in machine learning, either applied or theoretical. Specifically, this includes the following sub goals:

 

  1. Identify open research problems in machine learning or interesting applications.
  2. Pick a problem (either practical or theoretical) that interests you and propose a novel solution.
  3. Investigate the efficacy of your proposed solution.
  4. Write up your solution in a form that is adequate for submission to a conference in an appropriate area.

 

My hope is that everyone will submit a paper to a conference and/or journal.

 

Your Grade will be based on the novelty of your proposed solution, its analysis, the presentation of your research and related research in class, and the conference paper you produce.

 

Course Outline:

 

  • By January 31: Identify an area or application of machine learning that you are interested in.
  • By February 15: Identify an interesting open research problem in that area.
  • By March 15: Propose a solution.
  • By April 15: Investigate/modify your proposed solution.
  • By April 30: Write up your results in conference paper form.

 

Final conference paper due April 30. Emailed to me no later than Midnight!

 

 

You will be expected to do the following:

 

  • Present at least 4 papers that are related to your research topic in class.
  • Updates on your progress regularly in class.
  • Actively critique and comment while others are presenting in class.

 

I encourage you to work alone, although you may work in groups of 2 or 3. I will closely supervise every research project and will periodically present material in class that I believe is relevant.

 

 

Textbook: The Elements of Statistical Learning, by Hastie, Tibshirani, Friedman

 

 

 

Disability Accommodations

If you qualify for accommodations because of a disability, please submit to me a letter from Disability Services in a timely manner so that your needs may be addressed. Disability Services determines accommodations based on documented disabilities. (303-492-8671, Willard 322, http://www.colorado.edu/disabilityservices)

 

Religious Accommodations

If you feel you can't complete class work for religious reasons, please contact me as soon as possible to make arrangements.

 

Honor Code

The campus has adopted an Honor Code. It includes the following pledge which will be placed on all your exams and you will need to include on your assignments:

 

On my honor, as a University of Colorado at Boulder student, I have neither given nor received unauthorized assistance on this work.