Permanently Closed

This blog is now permanently closed.

I'm starting a new blog at http://www.andreweckford.com/blog/

IEEE North American School of Information Theory

The 2014 IEEE North American School of Information Theory (NASIT) will be held right here in T.O., from June 18-21, 2014. It will be held at the Fields Institute, on the lovely downtown campus of the University of Toronto.

This is a great event with a stellar lineup of speakers, and you should apply. Deadline is March 7.

Lab equipment for molecular communication

You may have seen our PLOS ONE paper about tabletop molecular communication, which received loads of media coverage. One of the goals of this paper was to show that anyone can do experiments in molecular communication, without any wet labs or expensive apparatus.

For molecular communication researchers, hobbyists, or anyone else who wants to do experiments like ours, my PhD student, Nariman Farsad, is making our apparatus available as an easy-to-use kit. (Full disclosure: the equipment is distributed through a small business founded by Nariman; I have no financial interest in the business.)

For further details and quotes, please contact Nariman at: nariman@cse.yorku.ca

[Updated] Doing the media rounds

My paper on molecular communication came out last night on PLOS ONE:

N. Farsad, W. Guo, and A. W. Eckford, "Tabletop Molecular Communication: Text Messages through Chemical Signals," PLoS ONE, vol. 8, no. 12, December 18, 2013.

Here I am giving an interview to CTV News Channel.

Here's coverage in the National Post.

I also did a radio interview, and a few other outlets picked up the story. I'll update with more links as I get them.

Updates:

• Video from the Globe and Mail
• News story on CHCH (Hamilton, Ontario)

Verdù interviews Fano

You must check out Sergio Verdù's amazing interview with Robert Fano in this month's IT Society newsletter. A giant of information theory, Fano was the third recipient of the Shannon award, and is probably best known in our community for the Fano inequality.

The sweep of Fano's career is incredible. He coined the term mutual information. At MIT in 1950, he gave what was likely the first course anywhere on information theory; it was in this class that David Huffman famously invented the Huffman code in a term paper.  But by the sixties, Fano was already out of information theory -- he wrote a book on electromagnetism, then got interested in computer science, becoming the founding director of the centre that would become CSAIL. Now 95 years old, he still keeps an office at MIT.

[Updated] Molecular Communication book: Radio interview on CKNW, Vancouver

I did a radio interview on The Shift with Mike Eckford, on CKNW in Vancouver, about my new book. Here it is on SoundCloud.

UPDATE: Here's an embedded version.
 

New book: Molecular Communication

How does it feel to (co-)write a book and hold the finished product in your hands? About like this:

Many, many thanks to my excellent co-authors, Tadashi Nakano and Tokuko Haraguchi, for their hard work; thanks to Cambridge for accepting this project and managing it well; and thanks to Satoshi Hiyama for writing a nice blurb.

Here's the back-cover text:

This comprehensive guide, by pioneers in the field, brings together, for the first time, everything a new researcher, graduate student or industry practitioner needs to get started in molecular communication. Written with accessibility in mind, it requires little background knowledge, and provides a detailed introduction to the relevant aspects of biology and information theory, as well as coverage of practical systems. The authors start by describing biological nanomachines, the basics of biological molecular communication and the microorganisms that use it. They then proceed to engineered molecular communication and the molecular communication paradigm, with mathematical models of various types of molecular communication and a description of the information and communication theory of molecular communication. Finally, the practical aspects of designing molecular communication systems are presented, including a review of the key applications. Ideal for engineers and biologists looking to get up to speed on the current practice in this growing field.

The book will be released on October 31 in North America. You can pre-order it on Amazon.

Happy trails, Voyager

Here's Voyager 2 looking back at Neptune and its moon Triton, possibly the most poignant of the amazing pictures it returned on its voyage of discovery. Its sister ship, Voyager 1, is now in the interstellar medium. [Image source.]

The issues of National Geographic with the Voyager pictures were probably the most dog-eared of my parents' collection. Back in undergrad, in the early days of the web, I remember realizing, "Hey, JPL has a website? You mean I can see the Voyager pictures whenever I want?!?!?!" Yes, I have always been a huge space nerd.

Voyager 1 is expected to keep transmitting for another ten years or so. Happy trails.

How close is the ISIT venue to the Istanbul protests?

Very likely it will all blow over by next month ... but if you're curious how close the ISIT 2013 venue is to the Istanbul protests in Taksim Square (about the possible development of Gezi Park), here's a quick graphic.

Indefinite hiatus

I may as well admit that the blog is on hiatus for the foreseeable future. I've got plenty of things to write about, but no time. Sabbaticals, they are busy! Expect this situation to continue until at least the fall.

Book on molecular communication

For the last 18 months, I've been working on a book on molecular communication. The final complete draft was just sent off for production. It should be printed later this year.

Many thanks to my excellent co-authors, Tadashi Nakano (Osaka University) and Tokuko Haraguchi (NICT); Tadashi in particular for doing most of the work, and for hosting my trips to Japan to hammer out the manuscript. Also many thanks to our contacts at Cambridge University Press for their help and support.

You can read a chapter here. Comments are welcome. (From: T. Nakano, A. W. Eckford, T. Haraguchi, Molecular Communication, Cambridge University Press.)

Just like the CCRAP party

In light of conference malware attacks, and conference phishing scams, and even the IEEE's crackdown on bad conferences, I'm going to suggest that IEEE SCAM is not the best possible acronym.

On Notation

I've spent the last week (and counting) trying to fix the notation in a long document that I've been writing, off and on, over the last 18 months. Since I wrote in fits and starts, I wasn't generally consistent with the notation, always thinking I could just fix it at the end. So when the end came, the results were predictably disastrous.

But I've had a long time to think/swear about notation, and I made some observations:

(Sorry for the image, it's much easier than trying to put LaTeX into Blogger.)

I used to think that there should be a notation stylebook, one notation to rule them all. Or that I would, once and for all, teach my graduate students the "right" notation. But now I don't think that's possible: there simply aren't enough degrees of freedom in a concise, readable notation to make it also rigorous and universally consistent. And you have to favor readability over universality, or what is a notation for?

Happy new year, happy new information theory blog

Via this month's IT Society newsletter, here's the Princeton-Stanford Information Theory b-log. Sergio Verdu is one of the authors.

I've already learned something useful, namely that I should have stopped using eqnarray years ago.

BIRS news

I'm delighted to report that my BIRS workshop proposal, entitled "Biological and Bio-Inspired Information Theory" (co-organized with Peter Thomas and Toby Berger), was accepted. The workshop will take place in October 2014.

BIRS is a gem of Canadian scientific research. My thanks to all the BIRS volunteers who make such a wonderful facility possible, especially the amazing Nassif Ghoussoub.

This Charming Man

As a fan of The Smiths from way back, I always thought Johnny Marr was the more talented member. After bouncing around several bands after The Smiths broke up, he'll be releasing his debut solo album early next year.

IEEE Dumps Bad Conferences from IEEE Xplore

This is interesting: the IEEE has a "Technical Program Integrity Committee" which has reviewed recent IEEE-sponsored conferences, looking for "inconsistencies in some conferences with regard to the quality of the peer review and technical program development." Some 160 conferences from 2010 and 2011 did not make the cut, and will now be barred from IEEE Xplore, the IEEE's digital library. This looks like the first action of the committee, which apparently started work in 2009.

First off, 160 bad conferences is a surprisingly large number. They say this is "less than 5%" of conferences from 2010 and 2011. I guess we can assume it's close to 5%, which still seems kind of high. In other words, out of every 20 conferences that the IEEE sponsors, one is bad.

Second, if you read their FAQ, it seems like they are concerned not only with quality, but with scope. For example:

Q - IEEE has notified us that the proceedings from this conference will not be included in IEEE Xplore® because they are out of scope. What is the scope of IEEE Xplore?
A - Content in IEEE Xplore must contain a significant electrical (and related) engineering component.

Q - How many out of scope papers can the conference proceedings have?
A - The conference proceedings should not have any out of scope papers.

Emphasis is added. So how is the committee judging this? There are several rejected conference titles with a life-science focus (an area in which I do some research), so were these rejected for scope or for quality? Where can we read this "scope"? What exactly are the rules surrounding interdisciplinary research in the IEEE?

Nate Silver's odd wager

Let's say you figured out a way to predict a coin toss. It's not a sure thing, but you think you can call the toss with 82% accuracy. Naturally you're excited about this, so you tell everybody.

The neighborhood bully gets wind of this, and confronts you. "That's stupid," he says. "Nobody can call a coin toss. It's fifty-fifty for everybody."

So you offer the bully a bet: you'll try to call one coin toss. If you call it right, you get $1000, if you're wrong, the money goes to him.

Wait, what? How does that decide anything?

Nate Silver is a brilliant statistician, and I think his call of the election is on the money. But his bet with Joe Scarborough solves nothing and confuses his point. Silver can be wrong and still win the bet with 50% probability, which Scarborough would be sure to point out. Worse, Silver can be completely right and still lose the bet with 18% probability. That's around one in six -- or the odds of rolling a one on a fair die. Not something I'd bet my reputation on.

If Silver wants to bet, surely he can craft a bet that he is sure to win if he is right, and sure to lose if he is not. This should be a case study in decision theory.

Would you include your blog in your T&P file?

What's the difference between a blog and a book, as far as "publications" are concerned? It's not quality -- it is not much harder to get a book published than it is to write the equivalent material in a blog. So if you are happy to list a book in your Tenure & Promotion file, would you also include your blog?

I know of people who have done it, like our former science librarian John Dupuis (his excellent blog is here). But I expect John is the exception rather than the rule.

I see a couple of things holding most researchers back from using their blog professionally.

On False Positives

Methods

One mature Atlantic Salmon (Salmo salar) participated in the fMRI study. The salmon was approximately 18 inches long, weighed 3.8 lbs, and was not alive at the time of scanning. The task administered to the salmon involved completing an open-ended mentalizing task. The salmon was shown a series of photographs depicting individuals in social situations with a specified emotional valence. The salmon was asked to determine what emotion the individual in the photo must have been experiencing. Stimuli were presented in a block design with each photo presented for 10 seconds followed by 12 seconds of rest. A total of 15 photos were displayed. Total scan time was 5.5 minutes.

 Source. See also.