Friday, October 28, 2016

Postdocs (Rabin and other) for this year


The Theory Group at Harvard has put out its ad for postdocs again this year.  It will be our second year for the Rabin Postdoc in theoretical computer science. "Rabin Fellows will be provided with the opportunity to pursue their research agenda in an intellectually vibrant environment with ample mentorship. While interaction with Harvard faculty, students, and visitors is encouraged, Rabin Fellows are free to pursue their own interests."  We also have open postdoc calls for the theory group, for the Center for Research on Computation and Society, and for the Institute for Applied and Computational Science.  More information on all these postdoc opportunities can be found here.  Candidates are encouraged to apply by December 1, although that's not a hard deadline.  Generally, we ask people what postdocs they're interested in, so you should only have to send in your materials once even if you're interested in more than one of these opportunities.


Tuesday, August 09, 2016

FOCS 16 Workshops/Tutorials Call for Proposals

I'm told that there's a call for proposals up for workshops/tutorials for FOCS 16.
Here's a link to the call for those who are interested....

Deadline is August 31, end of the month.

Wednesday, July 06, 2016

STOC Theory Fest 2017 (Montreal June 19-23)

As SIGACT chair I announced at the STOC’16 business meeting that starting in 2017, STOC will turn into a 5-day event, a Theory Fest. This idea was discussed at some length in a special session at FOCS 2014 and the business meeting at STOC 2015. Now the event is being planned by a small group (Sanjeev Arora, SIGACT ex-chair Paul Beame, Avrim Blum, and Ryan Williams; who also get guidance from me and STOC’17 PC chair Valerie King). We’re  setting up committees to oversee various aspects of the event.

Below is a brief announcement from the organizing group about their plans! Please feel free to write your suggestions here, though the outline below has been arrived at after a fair bit of deliberation and consultation.

Update:   Here is the web page for STOC 2017:  (http://acm-stoc.org/stoc2017/)

Michael.

------
We’re excited to bring you the Theory Fest in Montreal, June 19-23 2017!

The following are the major features (caveat: subject to tweaking in coming years):

(i)  STOC talks go into 3 parallel sessions instead of two.  Slight increase in number of accepts to 100-110.
(ii)   STOC papers also required to be presented in evening poster sessions (beer/snacks served).
(iii) About 9 hours of plenary sessions, which will include: (a) Three keynote 50-minute talks (usually prominent researchers from theory and nearby fields) (b)  Plenary 20-minute talks selected from the STOC program by the STOC PC ---best papers, and a few others. (c) Plenary 20-minute talks on notable papers from the broader theory world in the past year (including but not limited to FOCS, ICALP, SODA, CRYPTO, QIP, COMPLEXITY, SoCG, COLT, PODC, SPAA, KDD, SIGMOD/PODS, SIGMETRICS, WWW, ICML/NIPS), selected by a committee from a pool of nominations. (Many nominees may be invited instead to the poster session.)
(iv) 2-hour tutorials (three in parallel).
(v)    Some community-building activities, including grad student activities, networking, career advice, funding, recruiting, etc.
(vi)   A day of workshops; 3 running in parallel. (Total of 18 workshop-hours.)

Our hope is that workshop day(s) will over time develop into a separate eco-system of regular meetings and co-located conferences (short or long). In many other CS fields the workshop days generate as much energy as the main conference, and showcase innovative, edgy work.

Poster sessions have been largely missing at STOC, but they have advantages: (a) Attendees can quickly get a good idea of all the work presented at the conference (b) Grads and young researchers get more opportunities to present their work and to interact/network, fueled by beer and snacks. (c)Attendees get an easy way to make up for having missed a talk during the day, or to ask followup questions. (d) Posters on work from other theory conferences broadens the intellectual scope of STOC,

We invite other theory conferences to consider co-locating with the Theory Fest. To allow such coordination, in future the location/dates for the Theory Fest will be announced at least 18 months in advance, preferably 2 years. Even for 2017 it is not too late yet.  

Finally, we see the Theory Fest as a work in progress. Feedback from attendees will be actively sought and used to refashion the event.
  

Wednesday, June 01, 2016

STOC 2017 (and STOC 2016)

I'm starting to get asked about STOC 2017 dates -- we recently executed the hotel contract, and STOC will be June 18-23 in Montreal.  (It will really "start" the morning of the 19th, but we expect people will be getting to the hotel the 18th.)  Yes, that 5 days of STOC -- expect more announcements about how we'll be filling up those 5 days coming up.

And just a reminder -- STOC 2016 is around the corner, June 19-21, with a joint STOC/SoCG workshop and tutorial day on June 18.  The STOC 2016 page has more info, and even if you haven't signed up, you can still register as a walk-in on-site. 

Monday, May 30, 2016

Class Evaluations, Again

I've not been blogging much, but I have an excuse now that class evaluations have arrived.
  
By posting about the evaluations, the students know I read them.  And perhaps if any future students see this it will inspire them to try to make helpful comments, or at least fill out the evaluation.  This semester I taught the undergraduate Algorithms and Data Structures course, which at a bit over 170 students is the largest class I've taught at Harvard (and the largest the class has been since I've been here).  Let's see how I did on a class this big.

But before beginning, a word from our administration: 
Again this semester, some students may have accessed their course grades before completing their course evaluations. When this has happened in the past, we determined that the effect on evaluations was statistically undetectable, if any. Nonetheless, we wish to assure you that we understand that this semester's Q evaluations will not be 100% comparable to past terms and that the administration will take that fact into consideration when viewing the data.
If anyone can decipher that bit of bureaucratese for me,  please do.  (I admit, I found it amusing.)

Now, my favorite comment of the year:
Mitzenmacher can be a bit intimidating at times (I'm pretty sure he's the only professor under 40 years old whose class I've taken that I'm not on a first-name basis with)...
Whoever you are, you anonymous student you, it's wonderful that somehow you think I'm still under 40.  That was, well, a while ago...   (Also, you can call me Michael.  Mitzenmacher is hard say, and spell.  Maybe calling me Michael will make me seem less intimidating.)

A close second:


Take this course if you want to be good at CS. I also got a girlfriend out of it, so if you're looking for love, this is definitely the class you should take. Chicks dig algorithms.
Yes, they do.  And congratulations.  But seriously, please don't call them "chicks".  Some women find the term offensive, and I would like women in the class to feel welcome and comfortable.  It also makes you seem like you're living in the wrong decade.      

Another good one.  

Definitely the legend his former students make him out to be.
Just to clear, I interpret that as at best a neutral statement, not necessarily a positive one.  (The rest of the evaluation, I think, backs my interpretation.)  On an obviously more negative note: 

He is sarcastic and unhelpful. He even goes out of his way to be snarky and rude to students...
I appreciate the concern, but don't worry, I'm not going out of my way at all... the sarcasm and snark really just come naturally!

And apparently, it's rubbing off.  

This course could be improved by having longer and more difficult problem sets. Having tougher exams would be better too.
Hey, I don't think you've heard.  I'm supposed to be the sarcastic and snarky one. (The student's whole evaluation was like this.  Very funny and original -- surprisingly, nobody's ever done that before, that I can remember.)  At least a partial explanation for the evaluation above, perhaps:  
Piazza was riddled with dumb anonymous question (e.g., "Should my program handle negative ints?" "No, as stated in the assignment."). I think it should be course policy to delete these questions rather than give terse answers; each one causes hundreds of people to receive email notifications and spend tens of seconds digesting the poster's helplessness. I estimate that each poster is choosing to waste about two hours of our collective time rather than reread the assignment or email a TF.
Thank you for this comment;  it's a rare comment that addresses a challenging issue.  I continue to be ambivalent about Piazza.  It's very useful for students to be able to ask questions and get reasonably timely responses.  On the other hand, the ease of being able to ask questions leads to the issue you raised, as well as other similar problems.  (Students don't see many private questions of the form, "Is this an OK answer to this question...")  I'm not sure deleting the question is the correct response;  I can only imagine the reaction.  Already, people find my terse answers to some questions off-putting, or even "sarcastic" and "snarky"....
  
Now for some other useful comments.   
better handwriting would be nice.
Not the first time I've heard that complaint, although surprisingly it was rarely mentioned this year;  I did try to be better at the board.  But it's good to be reminded, frequently.    

Maybe integrate into Canvas or disregard it completely, it was weird having some stuff in one place and other stuff on the other.
Thank you for the useful comment.  Canvas is the tool we are being pushed at Harvard to use for course management, and for a class of this size, I felt forced to use it;  having students e-mail in assignments just wasn't going to be functional.  But Canvas really, really sucks, and caused at least as many problems as it solved (probably more).  I'd have liked more comments about Canvas, or really some advice on whether it's worth using at all.  

An odd comment: 
really sucks not having a good textbook for those of us who learn better on our own than in lecture;
Comments like this always make me wonder.  I do not require a textbook, but recommend both CLRS and Kleinberg-Tardos if students want one;  both books line up reasonably well with probably 2/3 of the course.  Did this student buy one of those?  I don't know.  I also find most things I teach have pretty good articles available on Wikipedia.  Many students in their comments appreciated the lectures (and the lecture notes), but for those who don't, there are certainly other resources available.   
  
Now some good responses to the question:  What did you learn?  How did this course change you?
Algorithms are fun! I'm good at theory!
Awesome!   
...I now view computer science as more than just coding. It is the art of problem-solving.
Awesome! 

Some truth in advertising:
Very useful course for anyone who wants to learn how to solve problems in the most computationally efficient way possible. That said, this course is not for everyone. Some will find it difficult.
Some, indeed.  Perhaps I can find another student who can clarify the above description.... 
This class is insanely hard (like really, really, really hard); don't let anyone tell you otherwise. I spent maybe 40-50 hours per week on the psets, and the tests (the midterm and final) were brutal.
To be fair, I don't think most students are spending 40-50 hours per week on the problem sets, and the self-reported numbers suggest not also.  But point taken, the class is hard.  (Some will find it difficult.)   

As usual, the overall summary is that the reviews contain a mix of positive and negative.  Let's end with some more positive ones, so I can think positive thoughts about the fact I'll be teaching this very large class again next year: 
Really interesting course material presented clearly. There was a nice balance of breadth and depth, and I liked the focus on randomized algorithms and the last couple lectures on topics tying the course together. Such a good class - thank you!
 No, no -- thank you!
Mitzenmacher is very good at responding quickly on Piazza, and often his responses are quite funny.
Thank you!  Some people apparently don't seem to get our sense of humor.  Same thing happens at home with the wife and kids -- they don't think I'm funny either.  
Professor Mitzenmacher is a great lecturer and provides lots of insight into algorithms. Problem sets are very well written, and most if not all problems are very thought-stimulating and are an apt level of difficulty. I also enjoyed the programming assignments -- it was great to physically code things and get a hands-on approach to the algorithms we learned.

For programming assignments 2 and 3, I would have hoped that Professor Mitzenmacher grade them as well--the TFs seem to be following a rubric and it is difficult for them to reward other inisghts and rather take off many points for more minor issues.
Thank you for the nice comments!  I'm glad you were happy with my grading of the first programming assignment, but you really want me grading all of them???  (Hmmm... actually, maybe I'll think about grading PA3....)
Lastly, professor Mitzenmacher is incredibly nice and fun to talk to if you get the chance.
Thank you!  But please, don't let the secret out....

Let's end with this summary, answering what you would want to tell future students:  
You probably have to take this. Cheer up! First of all, it's not nearly as bad as everyone says...
Thank you, I think?***



(***This last backhanded compliment reminds me of a joke from when I was a kid, quoting from https://en.wikipedia.org/wiki/F_Troop
Another running gag involves dialogue between Sergeant O'Rourke and Agarn. In many episodes O'Rourke says to Agarn, "I don't know why everyone says you're so dumb". After several lines of dialogue later, and occasionally after a commercial break, Agarn finally replies, "Who says I'm dumb?".
 )

Wednesday, May 25, 2016

SODA Information

Phil Klein passed on that there's a delay at SIAM in getting the SODA 2017 CFP up, and of course we want to get out the relevant information out to the community.  So he asked me to post the following:   

SODA 2017: The official SIAM symposium webpage is http://siam.org/meetings/da17/. This page does not yet have the call for papers. (My understanding is that the call for papers has yet to be approved by some SIAM staff who are out this week.) The deadlines are as follows: July 6 (short abstract and paper registration), July 13 (full submission).

The symposium will take place on January 16-19, 2017, in Barcelona, Spain.

For now, you can visit http://cs.brown.edu/~pnk/#soda17 for the basic information (deadlines, submission site, and program committee).

Wednesday, March 09, 2016

David Johnson

I was sad to learn that David Johnson passed away.  David was a leader in the theoretical computer science community for decades, both in his research and in his dedicated service to the community.  He is a role model, and should continue to be, especially for those who believe that theoretical computer science has a large role to play in the scientific world. 

Lance has already written a post describing David's long, inspiring list of accomplishments.  I encourage you all to read it. 

Monday, February 29, 2016

Distinguished Service Award

Nominations are open for the SIGACT Distinguished Service Prize.

More information is available at the SIGACT web site.  

Here's the key info:

Nominations

Nominations can be made by any member of the Theory of Computing community and should contain a statement of no more than 500 words explaining why the candidate deserves the award. The nomination can also include an additional separate listing of service activities, additional support letters, and other supplemental material such as a pointer to the candidate's CV. The nomination must include the name, postal address, phone number, and e-mail address of the nominator. Nominations are to be submitted electronically in PDF format by April 1, 2016, to the chair of the selection committee, Lane A. Hemaspaandra, at lane@cs.rochester.edu. Please put "SIGACT Distinguished Service Award Nomination" in the subject line.

Thursday, January 07, 2016

Graduating Bits Announcement

Boaz asked me to remind people about ITCS, and in particular the Graduating Bits Event.  Just in time for those of us reading application folders....

ITCS 2016 will be held at MIT this year Jan 14-16. As in past years, we will hold a "graduating bits" event where students and postdocs that looking for positions can give a short presentation about themselves and their work. This can be a great way to let people know what you've been up to. See http://www.boazbarak.org/itcs16bits/ for more details. 

If you are interested in participating, please send Boaz Barak ( b@boazbarak.org )  an email with the subject "Graduating bits" and the following information:

1) Name, Affiliation, status (student/postdoc)
2) Photo of yourself (web quality - no huge files please).
3) A short paragraph (3-4 sentences) about yourself and your research.
4) Homepage URL
5) Your presentation - 4 slides in PDF format, of which the first slide should be a title slide.

There is no deadline per se, but the presentations will be scheduled in the order of submissions so it’s “first comes first served”. I will also maintain a website with the photos and slides of all presenters, which should be a useful resource for anyone looking to hire theoretical computer scientists this year.