What is the reward for timely reviewing?

I know I complain about this all the time, but in my post-job-hunt effort to get back on top of things, I’ve been trying to manage my review stack.

It is unclear to me what the reward for submitting a review on time is. If you submit a review on time, the AE knows that you are a reliable reviewer and will ask you to review more things in the future. So you’ve just increased your reviewing load. This certainly doesn’t help you get your own work done, since you end up spending more time reviewing papers. Furthermore, there’s something disheartening about submitting a review and then a few months later getting BCC-ed on the editorial decision. Of course, reviewing can be its own reward; I’ve learned a lot from some papers. It struck me today that there’s no real incentive to get the review in on time. Parv and Anant may be on to something here (alternate link).

In the somewhat, but not too distant future…

I have accepted an offer to join the Toyota Technological Institute at Chicago this fall as a Research Assistant Professor. I’m excited by all of the opportunities there; it’s a great chance to dig deeper into some exciting research problems while learning a lot of new things from the great people they have there. It’s in Hyde Park, which is a great stepping stone for future career opportunities


Robert Tavernor, Smoot’s Ear : The Measure of Humanity – This is an interesting, albeit dry, history of measurement in Europe, starting from the Greeks and Romans, more or less, up through the development of the metric system. It’s chock full of interesting little facts and also highlights the real problems that arise when there is no standard as well as when trying to define a standard.

Naguib Mahfouz, Palace Walk – The first in Mahfouz’s masterpiece trilogy, this novel follows a very traditional family of an Egyptian merchant, who spends his time partying every night while terrorizing his family during the day. It’s set during the end of the British occupation at the end of WWI and the protests against the British that start at the end of the novel seem eerily relevant today.

Nell Irvin Painter, The History of White People – This is a popular history of theories of race, beauty, and intelligence and how they became entwined with skin color, head-shape, and other measurable quantities. It was an interesting read but felt a little incomplete somehow. Also, she uses the work “pride of place” too many times. It was distracting!

Vivek Borkar, Stochastic Approximation : a Dynamical Systems Viewpoint – This slim book gives a concise, albeit very technical, introduction to the basic methods and results in stochastic approximation. It’s fairly mathematically challenging, but because it’s to-the-point, I found it easier going than the book by Kushner and Yin.

The strong converse for the MAC, part 2

The last time we saw that a packing lemma, together with a good choice of reference output distribution r (the Fano-* distribution) could give a converse for nonstationary DMC:

\log M \le \sum_{t=1}^{n} I(X_t ; Z_t) + (\frac{ 2 }{1 - \lambda} 3 |\mathcal{X}| n)^{1/2} + \log \frac{2}{1 - \lambda}

where the distribution on (X^n,Z^n) is given by the Fano distribution on the codebook, i.e. uniform input distribution on codewords followed by applying the channel. The next step in the argument is to apply this result to the MAC with inputs X and Y.

Continue reading

A little puzzle

This came up as sub-problem during Young-Han’s group meeting today and we mulled over it for a few minutes but didn’t come up with a non-ugly answer. I’m sure, given the number of Real Mathematicians ™ who read this, that someone out there knows of an “obvious” explanation.

Suppose I give you p integers in an arbitrary order (where p is prime). While maintaining the order and using only addition, multiplication, and parenthesis, is it always possible to make an expression which evaluates to 0 mod p?

I think it’s true, but I’m sure there’s some special property of \mathbb{F}_p that I have forgotten. I guess further generalizations would include whether or not it’s possible for arbitrary p (not necessarily prime), how many elements of an arbitrary field you would need, and so on. I’d ask this on MathOverflow but… meh. It’s probably a homework problem.