Imaginary Potential

According to my Librarian, The National Geographic Channel will be re-airing a show on man made marvels tomorrow and next Thursday on antarctic construction, with a partial highlight of the IceCube Experiment.  IceCube is a cubic kilometer of arctic ice laden with photomultipliers, and a surface array to detect air showers.   The PMT’s pick up Cherenkov light from fast moving charged particles (mostly muons), which come from cosmic rays and neutrino collisions with the water molecules.  Once completed, you can think of IceCube as being, among other things, a giant cubic telescope that views the universe with neutrinos instead of light.  It also contains an earlier detector called AMANDA, which is shown here as the yellow cylinder.  Photo comes from  gallery.icecube.wisc.edu.

I went to the weekly computing seminar yesterday, because it was on a statistical data mining tool that is being simultaneously used by physics experiments and marketing firms. The speaker is a physics professor, used to work for the PLUTO Collaboration, DELPHI, and now is variously associated with CDF and CMS. The company, Phi-T, is now totally private, and employs a couple dozen ex-physicists, or physicists, depending on how much of a purist you are. The software is proprietary and closed source, and the speaker was severely vague about what specific tools were actually used, but there is an interface in C++/ROOT/C#/Lisp already made, so its (supposedly) trivial to use, with a discounted academic licence.

So what is it? Basically, you have a vector of measureables, like detector channels, and some target, like say Resonance mass. Or Age, profession, #kids, and your target is “How much will this person cost us in Health Care in the next n years.” You then train the thing on your historic or simulated data, and it generates Bayesian posteriori distributions for new data. This is pretty common in neural computing literature, but this thing seems actually practical.

The only really fascinating thing is the generality of the thing, which was (supposedly) applied with minimal expert consultation on problems like car insurance premiums, to B_s mixing at CDF. Here’s a list of referred journal articles with their stamp. So whats inside? A neural network you say? No! The guy said in most applications they skip the neural net entirely and just use “Other” statistical methods. It’s clear that he was using some kind of input decorrelation like principle component analysis, but he wouldn’t say what specifically. He used a bunch of phrases that were cryptic to me like “zero layer network” to mean something other than a perceptron (I asked), and “zero iteration training” of a network. Maybe these things mean something to yall statters, but nothing to me. Anyways, the output of whatever was a discretized probability histogram that got splined together.

I’m unconvinced that the “default settings” he mentioned could schedule re-stocks for the largest book distributor in Germany AND find the X(3872) resonance, but what do I know? He also said that the companies own stock were controlled by this thing, but that selling it for this purpose is somehow illegal. Anyone know what he was talking about? Here’s a paper on it, by the speaker.

In the end, the talk was a sales-pitch/head-hunt, but if anyone out there needs to solve a highly nonlinear problem and has a cushy grant, go nuts.

If you speak German, and like science history, you can download Weyl’s 1919 Zeit, Raum, Materie book in PDF for free from The Internet Archive. I wanted a sort of journalistic, “this is how it went down” kind of book, but that doesn’t seem to be the case. There’s some nice first person stuff in the forward, and some nice philisophical musing at the end, but mostly is seems textbook stylie. I just skimmed it a bit, so maybe the gems are buried deeper.

If anyone can find anything on his early attempt at gauge invariant therory of unified electronmagnetism and gravity, I’d be grateful to hear from you.

You can also check out a collection of original papers on Relativity from before 1920, too. These are all translated into english.

The undergraduates invited Stephen Wolfram last thursday to give the physics colloquium. It was very strange.

For those who don’t know, about five years or so ago Stephen Wolfram, the creator of Mathematica, released a ridiculously large book called “A New Kind of Science.” This book is entirely too large, so I have not read it (see also “The Bell Curve”). So I will have to rely on others who have read, and on his colloquium it to summarize what I understand are some of his points.

The first one is that complex phenomena can be emergent from incredibly simple laws/rules. I don’t think this is a point that anybody really argues with. In fact, I think a lot of current condensed matter theory relies on precisely this fact. In his talk, he spent a lot of time (way too much actually) emphasizing this point. His shtick is cellular automata, some of which can produce ridiculously pseudo-random looking patterns based on very simple rules for how the cells grow. So he spent a lot of time showing very pretty pictures of these automata, which was fine enough.

At some point, the talk took a very weird turn, and he started blatantly advertising the newest version of Mathematica. It really was strange, because the features he was advertising (in real time, on his computer), had seemingly nothing to do with the talk he was giving. Then he said something like: “for the past few years, we have been working on a new piece of software which will accomplish what nobody thought possible. It’s going to be huge and possibly change all of your lives.” He did not elaborate.

After this point, I began to sense frustration from the audience, as it had been forty five minutes, and the man hadn’t actually said anything. All the while, he had been promising to elaborate on his biggest and most suspect claim of the very large book: the simple rules of cellular automata can somehow generate fundamental laws of physics. Finally, with about ten minutes left, he made a series of grandiose claims without elaboration; somehow with cellular automata, the theories of special and general relativity can somehow be derived.

Now, I don’t doubt that Wolfram is a very smart man, smarter than I will ever be. He got his Ph.D. in physics at 20 or something, he invented mathematica. But if you are giving a talk to a physics audience and you have a claim like “GR can be derived from cellular automata,” well, you’d better spent your entire alloted time justifying that claim instead of inappropriately shilling for your software program. Because, if true (which I think I, and most real physicists are very skeptical about), it is ridiculously interesting.

The talk ended with a bit of awkwardness; Wolfram had gone over his time, but was very interested in taking questions from the audience. After a few of them, the undergraduate host tried to get him to stop. Wolfram pleaded for more time, but eventually the undergraduate just cut him off, nobody came out looking well from the exchange.

Maybe yall remember when I was writing some conference proceedings last month. I finished them at the last minute (like everything else I do) and mailed in the LaTeX source, the original figures in eps and the resultant dvi/ps/pdf, just like the webpage said. I saw something about instructions for a photograph-ready hardcopy, and assumed it was a courtesy for participants who work in caves* and code on pre-RISC DECs with orange-on-black CRTs. I just got an email from the secretary who is gathering the contributions for the publisher. Everyone has to send in hardcopies. Like, mail in paper. Like, tree-paper. Like, physically. Like, with a stamp. …? I could see it if you submit via Word or some other format that intentionally discourages interoperability, but I used their LaTeX format with no additional (ie modern) packages.

This wasn’t my first rodeo, and I ain’t never had to do this before. I asked an olderwiser about it, and they said “Sometimes they ask you for it. Don’t.” I mean, I will do it, but man, that’s silly.

*no offense to folks at neutrino/DM experiments. Those are mines, not caves.

If, today, you feel as if you should be doing work, but don’t really want to do any, may I suggest a video of a Sidney Coleman Lecture: Quantum Mechanics in Your Face? You’ll learn and be entertained.

The lecture is fantastic and consists of the late, great Coleman discussing a version of Bell’s theorem (which is much easier to understand than the standard treatment), and then going on to discuss the “mysterious” “collapse of the wavefunction”. It’s great stuff. All that’s required for enjoyment is a basic undergrad QM course…