It’s been one year since I started as executive officer (Caltech’s name for department chair) for our CMS department…and, not coincidentally, it’s been almost that long since my last blog post! But now, a year in, I’ve got my administrative legs under me and I think I can get back to posting at least semi-regularly.
As always, the first post back after a long gap is a news filled one, so here goes!
Caltech had an amazing faculty recruitment year last year! Caltech’s claim to fame in computer science has always been pioneering disruptive new fields at the interface of computing — quantum computing, dna computing, sparsity and compressed sensing, algorithmic game theory, … Well, this year we began an institute-wide initiative to redouble our efforts on this front and it yielded big rewards. We hired six new mid-career faculty at the interface of computer science! That is an enormous number for Caltech, where the whole place only has 300 faculty…
I’m a little late in posting about this, but at this point the schedule for Sigmetrics is up, as is the list of accepted papers at EC. So, it’s a good point to take a look at each to compare.
First of all, congrats to all the folks in RSRG with paper(s)…we managed to get one in at both conferences again this year, to extend our streak to 5 years in that regard! (Though my personal streak got broken this year with a gap on the EC side…) Here’re the two papers:
- Online Convex Optimization Using Predictions. Niangjun Chen, Anish Agarwal, Adam Wierman, Siddharth Barman, and Lachlan Andrew (Sigmetrics)
- Finding Any Nontrivial Coarse Correlated Equilibrium Is Hard. Siddharth Barman and Katrina Ligett (EC)
Second, it’s always interesting to look at the trends in the accepted papers at the two conferences. On the Sigmetrics side, there is a marked drop in papers related to data centers / cloud applications and also in papers related to energy. In both cases, this seems to be (at least partially) a consequence of the emergence of new top-tier conferences focusing on those areas with similar deadlines. I know this happened in my case… These papers have been replaced by the emergence of a class of papers at the intersection of control and learning (our paper is one such example) and the continued expansion of papers on markets & social networks. There are two sessions this year full of papers that could easily have been submitted to EC instead.
In contrast, topics in the EC submissions seem pretty stable. There is again a large focus on mechanism design and computational advertising models. One area that seems to have grown considerably though is papers related to online/sharing economies and networked marketplaces, with a number of papers focusing on uber/airbnb-type applications.
Happily, there is (for the first time I believe) one paper with energy markets as the focus (from Kesselheim, Kleinberg, Tardos)! I’ll be excited to take a look at that… So far, I haven’t taken the plunge and sent our work on that topic to EC (which is why I didn’t submit this year), but maybe it’s time to give it a shot.
I’ve posted before about the growing overlap in topics between the EC & Sigmetrics communities. Increasingly, the conferences are having sessions on very similar topics, but looking at the topics with very different perspectives. This was especially true for last year’s conferences. While some folks manage to cross the boundaries between these two communities, for the most part this has proven difficult (at least for my attempts)…
I think there’s a lot of value in finding ways to bring folks from these two communities together, which is one reason why I’m happy to announce NetEcon 2015, which I’m co-chairing with Patrick Loiseau and Aaron Roth. NetEcon is a workshop with a long history of bridging CS and Economics, and this year we’re hoping to take advantage of the fact that Sigmetrics and EC are co-located as part of FCRC to bring those two communities together as well.
To highlight this, we have three exciting keynotes lined up. One from Economics (Rakesh Vohra), one from the EC community (Eva Tardos), and one from the Sigmetrics community (R. Srikant). Additionally, to be consistent with the differing publication styles across the communities, we will allow accepted papers to have a 1-page abstract appear in the proceedings in order to ensure that the full version of the papers can be published in other venues without issue.
So, there’s no excuse not to send in your best on-going paper or work-in-progress!
In the last few months, the conferences I’ve attended have been on the extremes in terms of sizes. A few small targeted workshops (like NEGT, which I blogged about), and a few massive conferences (CDC & INFORMS). The contrast made me realize that I get very different things out of these two types of venues.
In particular, for me, there is a clear relationship between the number of talks I go to and the size of the conference.
There is a clear peak in the number of talks I attend around the size of a smallish single or double track conference… When there are 30-40 talks and 100-200 people, it seems that I end up spending most of my time actually in the talks, whereas in these massive mega-conferences like CDC and INFORMS, where there are 50+ parallel tracks, I end up spending almost no time in talks, and instead end up spending all of my time in the halls chatting/working with people. CDC this year was an extreme case for me — I ended up scheduling meetings with collaborators and prospective students/postdocs nearly all day every day, so I only managed to make it to ~1 talk per day! (Hence the lack of a “Report from CDC” blog post.)
(I wrote this during the workshop a few weeks ago, and just realized that I never actually hit “publish.” Better late then never, I guess!)
Every year in the fall, all the folks in southern California interested in the intersection of economics and engineering/computer science get together and have a two-day workshop that we call NEGT for “Network Economics and Game Theory.” Hosting duties rotate between USC, UCLA, and Caltech, and this year it was our job. The workshop is just wrapping up and, thanks to our amazing admin Sydney Garstang, everything went wonderfully!
There were lots of great talks, and the slides will eventually start to show up here. Of the many highlights, our two external speakers both gave really great talks. Our first keynote, Tim Roughgarden, gave a great overview of recent results in the area of approximate mechanism design. This is a direction that many folks in the Algorithmic Game Theory community have been pushing on in a while, but Tim showed some very interesting new results. Plus, it is always interesting to see how economists react to this direction, which is very different than the traditional viewpoint. Our second keynote, Markus Mobius, gave a really interesting empirical take on the power of social learning. He showed results from an experiment involving Harvard undergraduates performing a task that required social learning and was able to test various conjectures for how such learning occurs (as well as the magnitude of social learning that occurs). Given the huge focus in CS on models where we learn from our friends, it was quite interesting to see that the magnitude of such social learning is actually pretty small, and seems to occur only in vary specific ways.
The mathematics of planet earth is a joint initiative from a consortium of mathematical sciences organizations around the world (organized nominally by DIMACS) that has the goal of showcasing how mathematics can be useful in tackling our world’s problems. It started last year as a year-long focus, but has now expanded and will continue for the coming years as well. I’ve been to a few events organized under this program, but the reason for this post is to highlight the recent workshop on “Data-aware energy use” organized at UCSD a week or so ago.
All of the slides (and some videos) are now up from the Control and Dynamical Systems (CDS) 20th anniversary workshop that we held here at Caltech earlier this month.
The workshop was a huge success, and it was thrilling to see so many alums from the PhD program coming back. It’s really amazing to see how successful the program has been, and how varied the research is that they are doing now. CDS alums have become professors at all of the top 15 universities in the world (according to the London Times ranking), and they hold positions in a huge variety of departments: CS, EE, Control, Mech E, Math, and Bio. So, as a result, the applications covered during the CDS20 workshop were extremely broad: from bio and physiology, to communication networks and the smart grid, to machine learning and privacy.
And, not only was the worshop the 20th anniversary of CDS, it marked the kickoff our our new Computing and Mathematical Sciences (CMS) PhD program. We used the last session to highlight the evolution of CDS, which has resulted in the emergence of CMS. This new program is really modeled after the tenets of CDS that have proven so successful: seek rigor and relevance, and ensure that research is student-centric. I’ve written about this new program in a previous post, but I put together a slide deck for CDS20 that I think does a pretty good job of introducing the program. We can only hope that in 20 years we’ll have had as much impact as CDS…