Saturday, March 31, 2012

The Valiant Never Taste Death But Once

After reading this interesting excerpted article from Dick Teresi's book The Undead, which discusses the difficulties in defining death by a single, consistent set of criteria and the social qualms that stirs, I decided to check out the Amazon reviews. The associated ratings were (and still are) quite shockingly bad! They follow the classic "so bad it's good" distribution, with 5 5-star ratings, 1 3-star rating, and 33 1-star ratings. So, given that I am always up for a good controversy, I decided to read and review it myself. Ultimately I mostly side with the critics, giving it two stars. If you are interested in the subject matter, I'd suggest instead Kenneth Iserson's Death to Dust, which is a bit older but much more level-headed and thorough treatment of similar issues. 

Friday, March 30, 2012

What Makes Phrases Memorable

For 1000 movies, this study compared lines included on imdb's memorable quotes page to those that were not. People who hadn't seen the movies were able to pick the correct one 78% of the time, although, caveat lector, that's with only n = 68.

What features allow this above chance classification? The authors suggest 1) distinctiveness (i.e., a lower likelihood of coming from samples of standard English text), 2) generality (fewer personal pronouns, more present tense), and 3) complexity (words with more syllables and fewer coordinating conjunctions like "for" and "and").

Interestingly, their best support vector machine only correctly classified examples 64% of the time, so either the human data is somehow biased, or there are plenty more subtleties for machines to learn before they can best us humans in recognizing literary wit. 

Thursday, March 29, 2012

Indexing Wikipedia Article Submissions On Pubmed

I have complained before about few academics writing Wikipedia pages and instead writing reviews that few people will read. So, I feel compelled to admit that this is really cool:
We suggest a principal reason for this limited breadth and depth of coverage of topics in computational biology is one that affects a number of disciplines: reward. Authors in the biomedical sciences get academic reward for publishing papers in reputable journals that are indexed in PubMed and have associated digital object identifiers (DOIs).... 
Topic Pages are the version of record of a page to be posted to (the English version of) Wikipedia. In other words, PLoS Computational Biology publishes a version that is static, includes author attributions, and is indexed in PubMed. In addition, we intend to make the reviews and reviewer identities of Topic Pages available to our readership. Our hope is that the Wikipedia pages subsequently become living documents that will be updated and enhanced by the Wikipedia community...
I continue to be impressed by the innovation from the PLoS suite. 

Monday, March 26, 2012

Why Does Speed Variability Create Congestion?


Above are the results from one trial of an experiment designed to answer this question. Participants were randomly assigned to one of two groups, each with its own walking direction and color.

The authors defined "clusters" as groups of people walking in basically the same path, with some leeway. They then did simulations to determine the average lifetime of a cluster as a function of the group's variability in walking speed. As you can see, the greater the variability, the shorter the lifetime of the clusters.

N = the number of pedestrians in the simulation
This trend fits with their experimental results. Here's how the authors explain it:
[T]hose moving faster catch up with those walking slower, leaving an empty zone in front of the slow walkers ... [P]edestrians who are willing to walk faster than others make use of density gaps to overtake the slow walkers in front of them. By doing so, faster pedestrians move away from their lane, and meet the opposite flow head-on a few seconds later. This initial perturbation often triggers a complex sequence of avoidance maneuvers that results in the observed global instabilities. 
So here's a situation where more diversity, defined as inter-individual variability, leads to worse outcomes. Of course, as the authors mention, there are many other situations, such as collective decision making, where inter-individual variability is actually quite helpful.

Perhaps more diversity generally serves the function of pushing a group out of local optima. So you can think of diversity as shifting a group more towards the "explore" side of the exploration-exploitation trade-off. This would hurt in situations with a clearly defined goal, such as pedestrians walking in a circle as quickly as possible. But it might help in more complex situations.

Sunday, March 25, 2012

Comp Exams For Each Course

The solution I propose is comprehensive exams at the end of each course, much like Advanced Placement exams, that thoroughly and objectively distinguish students on merit alone. The emphasis in each classroom would then shift from fighting the teacher for high grades to cooperating with the teacher to learn the material necessary to perform on the exam.
That's from Andrew Knight, in an essay discussing problems that will not be new to anyone who is or has recently been in school; more here. This is exactly what I wanted during most of my science and math courses. The alternative is to place a greater emphasis on big standardized tests like the SAT, but there can be so much variability in results from just one day.

One question is whether such exams could be a part of classes that are less fact-based, such as history and english. There is actually a machine learning competition for automated essay grading going on right now. I don't pretend to know the answer to this question, but even if it is currently infeasible, that shouldn't stop the tests from being used in math, science, and foreign language classes.