Machine learning and learning theory research
Lev Reyzin points out the CI Fellows program is renewed. CI Fellows are essentially NSF funded computer science postdocs for universities and industry research labs. I’ve been lucky and happy to have Lev visit me for a year under last year’s program, so I strongly recommend participating if it suits you.
As with last year, the application timeline is very short, with everything due by May 23.
Most long conversations between academics seem to converge on the topic of reviewing where almost no one is happy. A basic question is: Should most people be happy?
The case against is straightforward. Anyone who watches the flow of papers realizes that most papers amount to little in the longer term. By it’s nature research is brutal, where the second-best method is worthless, and the second person to discover things typically gets no credit. If you think about this for a moment, it’s very different from most other human endeavors. The second best migrant laborer, construction worker, manager, conductor, quarterback, etc… all can manage quite well. If a reviewer has even a vaguely predictive sense of what’s important in the longer term, then most people submitting papers will be unhappy.
But this argument unravels, in my experience. Perhaps half of reviews are thoughtless or simply wrong with a small part being simply malicious. And yet, I’m sure that most reviewers genuinely believe they can predict what will and will not be useful in the longer term. This disparity is a lack of communication. When academics have conversations about reviewing, the presumption of participants in each conversation is that they all share about the same beliefs about what will be useful, and what will take off. Such conversations rarely go into specifics, because the specifics are boring in particular, technical, and because their is a real chance of disagreement on the specifics themselves.
When double blind reviewing was first being considered for ICML, I remember speaking about the experience in the Crypto community, where in my estimate the reviewing was both fairer and less happy. Many conferences in machine learning have shifted to doubleblind reviewing, and I think we have seen this come to pass here as well. Without double blind reviewing, it is common to have an “in” crowd who everyone respects and whose papers are virtually always accepted. These people are happy, and the rest have little voice. With double blind reviewing, everyone suffers substantial rejections.
We might say “fine, at least it’s fair”, but in my experience there is a real problem. From a viewpoint external to the community, when the reviewing is poor and the viewpoint of people in the community highly contradictory, nothing good happens. Outsiders (i.e. most people) viewing the acrimony choose some other way to solve problems, proposals don’t get funded, and the community itself tends to fracture. For example, in cryptography, TCC (not double blind) has started, presumably because the top theory people got tired of having their papers rejected at Crypto (double blind). From a process-of-research standpoint, this seems suboptimal, as different groups using different methods to solve similar problems are particularly the people who you would prefer talking to each other.
What seems to be lost with double blind reviewing is some amount of compassion, unfairly allocated. In a double blind system, any given paper is plausibly from someone you don’t know, and since most papers go nowhere, plausibly not going anywhere. Consequently, the bias starts “against” for all work, a disadvantage which can be quite difficult to overcome. Some time ago, I discussed how I thought motivation should be the responsibility of the reviewer. Aaron Hertzman strongly disagreed on the grounds that this belief could dead end your career as an author. I’ve come to appreciate his viewpoint to an extent. But, it misses the point slightly—the question of “What is good for the community?” differs from “What is good for the author?” In a healthy community, reviewers will actively understand why a piece of work is or is not important, filling in and extending the motivation as they consider the problem.
So, a question is: How can we get compassionate reviewing? (And in a fair way?) It might help somewhat for reviewers to actively consider, as part of their review, the level and mechanism of impact that a paper may have. Reducing reviewing load is certainly helpful, but it is not sufficient alone, because many people naturally interpret a reduced reviewing load as time to work on other things. And, some mechanisms seem to even harm. For example, the two-phase reviewing process that ICML currently uses might save 0.5 reviews/paper, while guaranteeing that for half of the papers, the deciding review is done hastily with no author feedback, a recipe for mistakes.
What creates a great deal of compassion? Public responsibility helps (witness workshops more interesting than conferences). A natural conversation helps (the current method of single round response tends to be very stilted). And time, of course, helps. What else?
For about 5 years, I’ve been the treasurer of the Association for Computational Learning, otherwise known as COLT, taking over from John Case before me. A transfer of duties to Phil Long is now about complete. This probably matters to almost no one, but I wanted to describe things a bit for those interested.
The immediate impetus for this decision was unhappiness over reviewing decisions at COLT 2009, one as an author and several as a member of the program committee. I seem to have disagreements fairly often about what is important work, partly because I’m focused on learning theory with practical implications, partly because I define learning theory more broadly than is typical amongst COLT members, and partly because COLT suffers a bit from insider-clique issues. The degree to which these issues come up varies substantially each year so last year is not predictive of this one. And, it’s important to understand that COLT remains healthy with these issues not nearly so bad as they were. Nevertheless, I would like to see them taken more actively into account than I’ve been able to persuade people so far.
After thinking about it for a few days before acting, I decided to go ahead with the transfer for another reason: I’ve been suffering from multitask poisoning. Partly this is Ada, but partly it’s many other things, each of which takes a small bit of my time, in aggregate leaving me disappointing people, myself in particular. The effect of this has been quite obvious in terms of the posting rate on hunch.net.
Fortunately, Phil Long was ready to take up the duties, and he’s well positioned to do so.
Despite the above, I found being treasurer not particularly difficult. The functions of the treasury part of ACL have been
After having been treasurer for a little while, I’m convinced that having a nonprofit to back a conference is a good idea easing many difficulties with relatively small effort.
An important aspect of the site is that it’s collaborative: by uploading just one program or dataset, a user taps into the entire network of existing programs and datasets for comparison. While data and code repositories do exist (e.g., UCI, mloss.org), MLcomp is unique in that data and code interact to produce analyzable results.
MLcomp is under active development. Currently, seven machine learn task types (classification, regression, collaborative filtering, sequence tagging, etc.) are supported, with hundreds of standard programs and datasets already online. We encourage you to browse the site and hopefully contribute more! Please send comments and feedback to mlcomp.support (AT) gmail.com.
At the PAC-Bayes workshop earlier this week, Olivier Catoni described a result that I hadn’t believed was possible: a deviation bound depending only on the variance of a random variable.
For people not familiar with deviation bounds, this may be hard to appreciate. Deviation bounds, are one of the core components for the foundations of machine learning theory, so developments here have a potential to alter our understanding of how to learn and what is learnable. My understanding is that the basic proof techniques started with Bernstein and have evolved into several variants specialized for various applications. All of the variants I knew had a dependence on the range, with some also having a dependence on the variance of an IID or martingale random variable. This one is the first I know of with a dependence on only the variance.
The basic idea is to use a biased estimator of the mean which is not influenced much by outliers. Then, a deviation bound can be proved by using the exponential moment method, with the sum of the bias and the deviation bounded. The use of a biased estimator is clearly necessary, because an unbiased empirical average is inherently unstable—which was precisely the reason I didn’t think this was possible.
Precisely how this is useful for machine learning isn’t clear yet, but it opens up possibilities. For example, it’s common to suffer from large ranges in exploration settings, such as contextual bandits or active learning.