| Contacts: | Erik Demaine and Oren Weimann. |
|---|---|
| When: | Typically on alternate Tuesdays, 5:30–6:30pm (with the option of staying longer). Sign up to the mailing list to learn precise meeting times. |
| Where: | Theory Lab, G575, in the Stata Center, MIT building 32. |
The idea of the open problems session is to allow students, faculty, visitors, etc. to get together and collaborate towards solving open research problems. The session is open to everyone, and the problems are usually aimed to a wide audience so that anyone who knows algorithms and/or mathematics at a reasonable level should at least be able to understand the problem, and ideally also be able to contribute. Some easier problems can be solved during a session; other problems will be solved after the session by people who were stimulated by the problem; and other problems we will find to be too difficult, but only an hour or two of time will have been invested. Once we solve a problem, a subset of the attendees can decide to participate in the writing up and publication of a research paper.
Problems will come from various fields relating (at least tangentially) to algorithms. Examples include problems in combinatorics, graphs, geometry, data structures, games, string matching, information retrieval, approximation, randomization, cryptography, and biology. Before each session, we will send an email announcing the session which includes the problem of the day, in case that influences attendance.
The problem session is a great way to get more minds thinking about your problem, and still be able to contribute and be involved in the solution. If you have any interesting problems you would like to discuss in a session, please email Oren Weimann. In particular, visitors (e.g., giving theory seminars) are invited to submit a problem and lead a session during their visit. We will try to accomodate requests, but it may be a while before we can get to your problem. Meanwhile, your problem may be listed on our private webpage
For a list of several open problems, particularly in computational geometry, see The Open Problems Project.
To prevent forgotten results, during each problem session an attendee will be designated as scribe who prepares electronic notes of what was discussed. These notes will be posted on a password-protected webpage available to attendees of the session. For privacy, these notes are not public except in special circumstances. Please contact the organizers before distributing notes. Don't worry; anything noteworthy will eventually turn into a publication or technical report.
Attendance is open to anyone and visitors are always welcome. However, it would be good if you first email Oren, so that we can introduce you at the session. We have a mailing list, which carries a relatively small amount of traffic about upcoming problem sessions, what problems we plan to discuss, and updates on previous problems.
The MIT algorithmic open problem session was founded in October 2005 by Erik Demaine and Oren Weimann. Smaller-scale MIT open problem sessions were run by Erik Demaine in the context of two classes, 6.897: Advanced Data Structures (Spring 2005, with Mihai Pǎtraşcu) and 6.885: Folding and Unfolding in Computational Geometry (Fall 2004). In many ways these sessions follow the University of Waterloo Algorithmic Problem Session, which was cofounded by Therese Biedl and Erik Demaine in August 1999, and is currently organized by Therese Biedl and Alex Lopez-Ortiz. Both problem sessions were inspired by similar meetings around the world, in particular the Algorithms Seminar at SUNY Stony Brook. The University of Waterloo session has also inspired other similar meetings including the Computer Science Research Lunch at Smith College, and the University of Waterloo Bioinformatics Problem Session.
