Machine Learning and Artificial Intelligence Reading Group
We are the machine learning and artificial intelligence reading group (AIRG) in the Computer Sciences Department at the University of Wisconsin – Madison. We gather weekly during the semester to discuss papers in the areas of machine learning (ML), artificial intelligence (AI), natural language processing (NLP), bioinformatics, and computer vision (CV). We are the primary reading group for the AI area and cater to all its various topics, but mostly focus on ML.
AIRG meets Wednesdays from 4-5pm in CS 4310 during spring 2018.
Occasionally, when appropriate, AIRG is canceled in favor of the Statistics Seminar.
The mission of AIRG is to provide a forum to
- Learn about AI and ML
- Improve ability to read and understand academic literature
- Practice critical analysis and discussion of research
We strive to accommodate differing levels of education and experience and to provide a welcoming, accessible, open, and safe environment for all participants.
All UW students and interested members of the public are welcome to participate in AIRG.
- It’s casual. No pressure!
- Learn about interesting and exciting science with like-minded people
- Build your academic skills (reading, analysis / criticism, writing, presenting)
- Get faster at reading papers
- Keep up with developments in ML and AI
- Meet your colleagues
- Satisfy your curiosity and explore new topics
- Build your résumé (presentations go down in history!)
- Earn the respect of your peers
How it Works
Each week a member  leads the discussion of a paper of their choice. The member volunteers to present a paper (or other material, see below) by contacting the coordinators. About a week before the assigned meeting date, the member announces  their paper via the mailing list. Other members then read the paper  and come to the meeting prepared to discuss it. The volunteering member presents an overview of the paper (about 30-45 minutes if uninterrupted) and leads the ensuing discussion. Questions of any kind are always welcome at any time.
 A member is anyone who participates!
 A good announcement includes a brief, motivating summary of the paper (usually not the abstract), a publicly-accessible permalink (or the PDF if a permalink to the paper is not available), and a reminder of the meeting time and place (especially if a “special session”).
 Read the paper to the best of your ability and availability. Skim it if you need to. Read the abstract. Glance at the figures. Just show up!
How to Participate
- Be interested in AI and ML
- Be willing to delve into academic literature
- Show up to meetings (as often as you can!)
- Read papers (as best you can!)
- Lead paper discussions (optional but beneficial)
- Join the mailing list (optional but essential):
AIRG. The mailing
list is the primary means of communication. All readings are
announced via the mailing list. The list is limited to those with a
*wisc.edue-mail address or members of the public that have shown up to a meeting and asked to sign up.
Ideas for What to Present
- Papers from conferences and journals (e.g. NIPS, ICML, JMLR, KDD, IJCAI, AAAI, IJCAI, MLJ, UAI, ECML-PKDD, AISTATS, RECOMB, CVPR)
- Papers / Chapters from the PhD qualifying exam list. This material is always worth reviewing, even for advanced students.
- Papers / Chapters you have read as part of a class (especially appropriate for newer students to lead)
- Papers you have read for your research
- Material of personal interest and relevance to AIRG
- Slides about a paper from a talk given at a conference
- Practice talks of relevance to AIRG
The material should be short enough to discuss in an hour, although, for longer topics, we have occasionally arranged “special sessions” that extend the meeting time or span multiple meetings.
- Aubrey Barnard, the Monarch of AIRG
- Ronak Mehta, the Hand of the Monarch
We have an archive (under construction) of all the AIRG meetings going back to 2002!