Skip to content


Harvard Center for Research on Computation and Society: Call for Fellows and Visiting Scholars

The Harvard Center for Research on Computation and Society (CRCS) solicits applications for its Postdoctoral Fellows and Visiting Scholars Programs for the 2015-2016 academic year. Postdoctoral Fellows receive an annual salary of approximately $63,000 for one year (with the possibility of renewal) to engage in a program of original research, and are provided with additional funds for travel and research support. Visiting Scholars ordinarily come with their own support, but CRCS can occasionally offer supplemental funding.

We seek researchers who wish to interact with both computer scientists and colleagues from other disciplines, and have a demonstrated interest in connecting their research agenda with societal issues. We are particularly interested in candidates with interests in:

· Economics and Computer Science
· Health Care Informatics
· Privacy & Security
· Technology & Accessibility
· Automation & Reproducibility of Data Analysis
The ideal researcher will have a technical background in an area related to computer science, and a desire to creatively use those skills to address problems of societal importance. CRCS is a highly collaborative environment and we expect Fellows and Scholars to engage with researchers both inside and outside of computer science.

Examples of projects may be found at http://crcs.seas.harvard.edu/research

There are numerous opportunities for CRCS Fellows and Visiting Scholars to engage with Harvard faculty, students, and scholars in computer science and other disciplines, including the bi-weekly CRCS Lunch Seminar series, various informal CRCS lunches, and other research group meetings. Additionally, CRCS has close ties with Harvard’s Berkman Center for Internet and Society, and CRCS Fellows attend the weekly Berkman Fellows’ meeting.

Harvard University is an Affirmative Action/Equal Opportunity Employer. We are particularly interested in attracting women and underrepresented groups to participate in CRCS. For further information about the Center and its activities, see http://crcs.seas.harvard.edu/.

Application Procedure

A cover letter, CV, research statement, copies of up to three research papers, and up to three letters of reference should be sent to:

Postdoctoral Fellows and Visiting Scholars Programs
Center for Research on Computation and Society
crcs-apply@seas.harvard.edu

The cover letter should describe what appeals to you about joining CRCS and describe how you would connect with the existing community. Please also make clear in your cover letter whether you are applying for the Postdoctoral Fellow or Visiting Scholar position, as well as whether you are supplied with your own funding.

References for Postdoctoral Fellows should send their letters directly, and Visiting Scholar applicants may provide a list of references rather than having letters sent.

The application deadline for full consideration is Monday, December 1, 2014.

Posted in Uncategorized.


Which research results will generalize?

One approach to AI research is to work directly on applications that matter — say, trying to improve production systems for speech recognition or medical imaging. But most research, even in applied fields like computer vision, is done on highly simplified proxies for the real world. Progress on object recognition benchmarks — from toy-ish ones like MNIST, NORB, and Caltech101, to complex and challenging ones like ImageNet and Pascal VOC — isn’t valuable in its own right, but only insofar as it yields insights that help us design better systems for real applications.

So it’s natural to ask: which research results will generalize to new situations?

Continued…

Posted in Machine Learning.


Prior knowledge and overfitting

When we talk about priors and regularization, we often motivate them in terms of “incorporating knowledge” or “preventing overfitting.” In a sense, the two are equivalent: any prior or regularizer must favor certain explanations relative to others, so favoring one explanation is equivalent to punishing others. But I’ll argue that these are two very different phenomena, and it’s useful to know which one is going on. Continued…

Posted in Uncategorized.


ICML Highlight: Fast Dropout Training

In this post, I’ll summarize one of my favorite papers from ICML 2013: Fast Dropout Training, by Sida Wang and Christopher Manning. This paper derives an analytic approximation to dropout, a randomized regularization method recently proposed for training deep nets that has allowed big improvements in predictive accuracy.   Their approximation gives a roughly 10-times speedup under certain conditions.  Much more interestingly, the authors also show strong connections to existing regularization methods, shedding light on why dropout works so well. Continued…

Posted in Machine Learning, Recent work.


Testing MCMC code, part 2: integration tests

This is the second of two posts based on a testing tutorial I’m writing with David Duvenaud.

In my last post, I talked about checking the MCMC updates using unit tests. Most of the mistakes I’ve caught in my own code were ones I caught with unit tests. (Naturally, I have no idea about the ones I haven’t caught.) But no matter how thoroughly we unit test, there are still subtle bugs that slip through the cracks. Integration testing is a more global approach, and tests the overall behavior of the software, which depends on the interaction of multiple components. Continued…

Posted in Uncategorized.


Compressing genomes

Here’s an interesting question: how much space would it take to store the genomes of everyone in the world? Well, there are about 3 billion base pairs in a genome, and at 2 bits per base (4 choices), we have 6 billion bits or about 750 MB (say we are only storing one copy of each chromosome). Multiply this by 7 billion people and we have about 4800 petabytes. Ouch! But we can do a lot better. Continued…

Posted in Compression.


Testing MCMC code, part 1: unit tests

This post is taken from a tutorial I am writing with David Duvenaud.

Overview

When you write a nontrivial piece of software, how often do you get it completely correct on the first try?  When you implement a machine learning algorithm, how thorough are your tests?  If your answers are “rarely” and “not very,” stop and think about the implications.

There’s a large literature on testing the convergence of optimization algorithms and MCMC samplers, but I want to talk about a more basic problem here: how to test if your code correctly implements the mathematical specification of an algorithm. Continued…

Posted in Computation, Machine Learning.


The Central Limit Theorem

The proof and intuition presented here come from this excellent writeup by Yuval Filmus, which in turn draws upon ideas in this book by Fumio Hiai and Denes Petz. Suppose that we have a sequence of real-valued random variables

(1)   \begin{equation*} X_1, X_2, \ldots . \end{equation*}

Define the random variable

(2)   \begin{equation*} A_N = \frac{X_1 + \cdots + X_N}{\sqrt{N}} \end{equation*}

to be a scaled sum of the first N variables in the sequence. Now, we would like to make interesting statements about the sequence

(3)   \begin{equation*} A_1, A_2, \ldots . \end{equation*}

Continued…

Posted in Probability, Statistics.

Tagged with , .


JIT compilation in MATLAB

A few years ago MATLAB introduced a Just-In-Time (JIT) accelerator under the hood. Because the JIT acceleration runs behind the scenes, it is easy to miss (in fact, MathWorks seems to intentionally hide it so that users do not change their coding style, probably because the JIT accelerator is changed regularly). I just wanted to briefly mention what a JIT accelerator is and what it does in MATLAB. Continued…

Posted in Computation.


Introspection in AI

I’ve recently come across a fascinating blog post by Cambridge mathematician Tim Gowers. He and computational linguist Mohan Ganesalingam built a sort of automated mathematician which does the kind of “routine” mathematical proofs that mathematicians can do without backtracking. Their system was based on a formal theory of the semantics of mathematical language, together with introspection into how they solved problems. In other words, they worked through lots of simple examples and checked that their AI could solve the problems in a way that was cognitively plausible. The goal wasn’t to build a useful system (standard theorem provers are way more powerful), but to provide insight into our problem solving process. This post reminded me that, while our field has long moved away from this style of research, I think there’s still a lot to be gained from it. Continued…

Posted in Machine Learning, Meta.