George and the NIE : the argument from information theory

George makes the following argument: let X be a binary random variable that equals 1 if Iran has an active nuclear weapons program and 0 if not. Suppose that last month we knew that P(X = 1) = p > 1/2. Then we can measure our uncertainty about X via its entropy:

H(X) = hb(p) = – p log p – (1-p) log (1-p)

Here hb(p) is the binary entropy function. Now let Y be a random variable representing the NIE. We know that conditioning reduces entropy:

H(X | Y) ≤ H(X)

Let p’ be our new probability that X = 1 conditioned on the evidence Y. We cannot have p’ < p, because then hb(p’) > hb(p), which is a contradiction. Therefore p’ ≥ p and therefore the NIE shows that the chance Iran has an active nuclear program is even higher than before.

Exercise: Explain the error(s) in George’s argument.

Extra Credit: Write a short essay explaining why one should not abuse information theory for political ends.

unlocked networks

KQED‘s Forum program had an hour on Verizon’s new decision to unlock its network and the future of consumer wireless technologies. The first part is more about business and consumers, but the conversation wends its way later to issues of unloading cellular traffic to WiFi networks (much as Blackberries do now, I guess). The whole “femtocell” idea is an interesting one, as is the mesh network that Meraki and others are proposing. Of course, the hardest part is convincing consumers to adopt/use the changes, but there are plenty of research questions in there as well, and even good theory questions.

As an added bonus, a caller towards the end named Rajiv who is a “researcher in mobile internet” called in to say that these changes are not really “new innovation.” At the time I thought, “maybe that’s Rajiv Laroia,” but I doubt it.

Adjuncts in the academy

The tone of this NY Times article on the use of adjuncts struck me as odd. Perhaps because I am inside the ivory tower, the handwringing over the quality of education provided by adjuncts seems over-exaggerated. A bigger problem seems to be how the adjuncts themselves have no power and no real support, a point which is mentioned in the article but with a different spin. Some of this came out in the letters to the editor, but I think the article’s bias shows a bit about who the NY Times’ audience is — parents worried about whether they’re getting a good deal for their money on their kid’s education.

Mechanism Design

I read this morning that Maskin, Myerson, and Hurwicz won the Nobel for their work on mechanism design. I’ve gotten interested in auction theory in the last year, so for once I actually understand what the economics prize is being given for before reading all about it in the papers.

Hello Kitty : the mark of shame

Thai police use Hello Kitty armbands as punishment:

The armband is large, bright pink and has a Hello Kitty motif with two hearts embroidered on it.

So if you come late to work at the station — you have to wear Hello Kitty! I know so many people for whom this would not be considered punishment that it just seems plain weird to me. Of course, the arm band probably clashes horribly with the uniform color…

Prof. Sergio Servetto dies in plane crash

Professor Sergio Servetto died on Tuesday in a tragic plane crash. It’s a real shock, and great loss. I had just talked to him at ISIT in Nice. He was very active in getting the IT student society going and will be missed greatly. I was unsure if I should link to the news here, but I think that there are people who read this blog, knew him, and would want to know.

UPDATE : The IT Student Society has more information.

Jerry Hadley has died

Tenor Jerry Hadley has died a week after an apparent suicide attempt. I saw him twice that I remember — once in recital, which was very inspiring, and once in Harbison’s The Great Gatsby at the Met, where he was good but a bit impenetrable. It’s a great loss — he was one of the great American tenors of the last few decades.