APIs for hardware : architecting cellphones and networks

We had two talks by here at UCSD on Tuesday. The first was Cell Phones: How Power Consumption Determines Functionality by Arvind, and the second was Software-Defined Networks by Nick McKeown. These talks had a lot in common: they were both about shifting paradigms for designers, and about approaching the architecture of hardware from a software point of view.

Continue reading


More content-ful posts to come soon, I swear. I got sidetracked by job applications. ‘Tis the season, you know…

The REAL STORY of Alice and Bob. Classic investigative journalism (h/t Bikash Dey)

Tyler Perry’s For Colored Girls gets panned. I should mention that the play was done at MIT my first year there and I still remember it as one of the most affecting pieces of theater that I saw during my time there (and maybe after as well).

Scott McLemee on the poverty of the Rally to Restore Sanity.

Winners of the SD Asian Film Festival “Interpretations” contest. Like contentless scenes from your acting/directing class, but with film!

SAALT put out a report called From Macacas to Turban Toppers: The Rise in Xenophobic and Racist Rhetoric in American Political Discourse (PDF). Reading it is keeping me up past my bedtime. (via Sepia Mutiny).


The Hakawati (Rabih Alameddine) — an expansive novel, framed by the story of a son who has gone to the US coming back to visit his dying father in Lebanon. The sharply drawn tension and anguish of the present shifts rapidly through old family stories, to the story of Baybars (and parts in between). It’s hard to pick up the strands initially, but it’s a rewarding read once you get into it.

A Short History of the American Stomach (Frederick Kaufman) — a quick read, repeats of some stories from Harpers I had read. It might appeal to people who like Sarah Vowell’s writing, but it’s too heavy on snark for me. Good for picking up some cocktail-hour conversation pieces, if you enjoy talking about the puking habits of Puritans at cocktail hours.

The Magicians (Lev Grossman) — I enjoyed this book, even though some people call it Hipsters in Narnia. It is a bit of that, but I couldn’t really put it down (= brain candy). Recommended for those who want a jaded view of Harry Potter.

Ghostwritten (David Mitchell) — I read this one after reading Cloud Atlas, which I absolutely loved. It’s written in a similar style, with interlocking stories, but more direct storytelling going on than, say, if on a winter’s night a traveler. Maybe I just like the relay-race novel. In any case, definitely engrossing, if a bit… bleak? It’s simultaneously lush (descriptively) and bleak (psychologically).

Gaming the Vote (William F. Poundstone) — a popular nonfiction book about elections, the spoiler effect, and the history of voting systems. It’s larded with examples of elections from US history and makes for an engrossing read. Most of the focus is on the weaknesses of first-past-the-post and other methods of determining winners, but it’s a nice accessible read.

Nixing negative reviewers

A question came up while chatting with a friend — how do you tell the editors of the journal to not ask certain people for a review? Say you submit a paper to a journal and in the cover letter you want some language to the effect that “please don’t choose Dr. X as a reviewer, since they will be biased.” This must be a relatively common situation, especially where people have axes to grind, and what better way to grind them than while reviewing the other camp’s paper or grant proposal?

Let’s create a cartoon situation: suppose Dr. X really hates your guts (intellectually, of course) — this is actually the case, and not just your own misperceptions of Dr. X. I know that at some schools for tenure cases the candidate can give a list of people not to ask for letters. But in the context of paper submission, hows can you politely suggest that Dr. X may not be the most objective reviewer for your paper?


The fog in San Francisco (h/t Erin Rhode).

A general approach to privacy/utility tradeoffs, with metric spaces! A new preprint/note by Robert Kleinberg and Katrina Ligett.

Max breaks it down for you on how to use the divergence to get tight convergence for Markov chains.

The San Diego Asian Film Festival starts on Thursday!

Apparently China = SF Chinatown. Who knew? Maybe the fog confused them.

Mean absolute deviations for binomial random variables

Update: thanks to Yihong Wu for pointing out a typo in the statement of the result, which then took me months to get around to fixing.

I came across this paper in the Annals of Probability:

Mean absolute deviations of sample means and minimally concentrated binomials
Lutz Mattner

It contains the following cute lemma, which I didn’t know about before. Let S_n have binomial distribution with parameters (n,p). Let b(n,k,p) = \mathbb{P}(S_n = k). The first two parts of the lemma are given below.

Lemma. We have the following:

  1. k_0 \in \arg\max_{k \in \{0,\ldots, n\}} b(n,k,p) if and only if k_0 \in \{0,\ldots, n\} and (n+1)p - 1 \le k_0 \le (n+1) p.
  2. (De Moivre’s mean absolute deviation equality) \sum_{k=0}^{n} |k - np| b(n,k,p) = 2 v (1-p) \max_{k \in \{0,\ldots, n-1\}} b(n-1,k,p), where v is the unique integer between np < v \le np + 1.

The second part, which was new to me (perhaps I’ve been too sheltered), is also in a lovely paper from 1991 by Persi Diaconis and Sandy Zabell : “Closed Form Summation for Classical Distributions: Variations on a Theme of De Moivre,” in Statistical Science. Note that the sum in the second part is nothing more than \mathbb{E}|S_n - np|. Using this result, De Moivre proved that \lim_{n \to \infty} \mathbb{E}|S_n/n - p| \to 0, which implied (to him) that

if after taking a great number of experiments, it should be perceived that the happenings and failings have been nearly in a certain proportion, such as of 2 to 1, it may safely be concluded that the probabilities of happening or failing at any one time assigned will be very near in that proportion, and that the greater the number of experiments has been, so much nearer the truth will the conjectures be that are derived from them.

Diaconis and Zabell show the origins of this lemma, which leads to De Moivre’s result on the normal approximation to the binomial distribution. As for the proof of the L_1 convergence, they call the proof in the p = 1/2 case “simple but clever, impressive if only because of the notational infirmities of the day.” De Moivre’s proof was in Latin, but you can read a translation in their paper. A simple proof for rational p was given by Todhunter in 1865.

For those with an interest in probability with a dash of history to go along, the paper is a fun read.