Information theory, emotion, music

From The Information Theory of Emotions of Musical Chords, posted on ArXiV today:

My basic hypothesis is as follows. While perceiving a separate musical chord, the value of a relative
objective function L that is directly related to the main proportion of pitches of its constituent
sounds is generated in the mind of the subject. The idea of an increasing value of the objective
function ( L>1 ) accompanied by positive utilitarian emotions corresponds to a major chord, while
the idea of a decreasing value of the objective function ( L<1 ) accompanied by negative utilitarian
emotions corresponds to a minor chord.

I naturally thought of The Mathematics of Love:


Posting has been light

Due to extensive traveling around. At the end of August and beginning of September I attended a workshop at the American Institute of Mathematics on Permanents and modeling probability distributions. It was a lot of fun, and a lot different than any workshop I’d been to before. There were only 20 or so participants and every day we’d break out in smaller groups to actually work on some sub-problem or area. By the end of the week I felt like I’d had a crash course in large-alphabet probability estimation and also got a sense of the huge range of problems and techniques (from exchangeable random partition processes to Markov chain Monte Carlo) involved in tackling them.

Last week I gave a talk on gossip algorithms at UC Davis and got a chance to talk with a lot of people there. It was a great trip, except for the 6:30 flight time out of San Diego. Then last weekend there were heirloom tomatoes:

Heirloom Tomato with bread and Cowgirl Creamery Red Tail cheese

Heirloom Tomato with bread and Cowgirl Creamery Red Tail cheese

And also peach pie:
Peach pie...

Peach pie...

Mmmm, pie.

Criticism of open access is backwards, as usual

Inside Higher Ed has a piece today on the presidents of liberal arts colleges writing to support the Federal Research Access Act of 2009. The law would make federal agencies that sponsor research come up with methods for archiving and publishing research that they fund so it would be “made immediately available” to the public. It would (essentially) apply only to journal papers, which raises a question about computer science, which lives the fast-and-dangerous conference life.

The article ends with a reaction from “Martin Frank, executive director of the American Physiological Society and coordinator of the Washington D.C. Principles for Free Access to Science,” who claimed that the since there are many foreign journal subscribers, the argument that taxpayers should have access to the research is not very strong. Frank is concerned with non-profit publishers (such as professional societies like the IEEE), but in his eagerness to protect his own turf he completely ignores the fact that mega-publishers like Elsevier and the Nature Publishing Group are based in other countries. Elsevier is Headquartered in Amsterdam and NPG is run by Macmillan, which “is itself owned by German-based, family run company Verlagsgruppe Georg von Holtzbrinck GmbH.”

If Mr. Frank wants to make a nativist argument against an open access mandate, then perhaps he should support a ban on wasting American taxpayer dollars to fund foreign publishing houses. The whole “taxpayer” argument in the end is marketing for both sides — although in principle any citizen should have access to government-funded research, the real volume comes from universities and industry. Federal money is used many times over for the same piece of research — once to fund it and then once for every (public) university library which has to buy a subscription to the journal where the result was published. University libraries will not stop subscribing to the IEEE journals just because the NSF and DARPA funded research will be made available in (probably separate) repositories run by the NSF and DARPA. If a non-profit is publishing its journals at cost then they should still be affordable. The for-profit publishers are the ones who will have to realize that the “value added” by the Nature brand is not worth the markup they charge.