Saturday, February 28, 2009

University of Illinois at Urbana-Champaign

I just got back from UIUC in order to see if that is someplace I would want to go for graduate school. It was pretty exciting.

On Thursday, we got there and had a "social event" where we got to have pizza, play games that they had put out (like Texas Hold-Em and Stratego) and talk to each other. There were several professors there too. One of the other prospective graduate students brought along his wife, who was applying for a Masters in public policy, and obviously just stood there in confusion when we were talking about all our computer science stuff. We also got free rooms at the local Hampton Inn. (It's too bad that we didn't get to stay at the hotel a few blocks away, because then she could have said, "I may not be a computer science student, but I did stay at a Holiday Inn Express last night.")

On Friday we had the majority of the events. I got to see lots of cool demonstrations, such as a projector that could project onto a 3 dimensional surface. By this I mean that there is a white box (or some other shape) in the middle of the room, and the projector senses the shape of the box and alters its projection to conform to the shape and position of the box. I also got to see the robotics lab, where they had a course in the middle of the room that they had modified Roombas to navigate around. Finally, I got to try out the graphics lab's "head scanner" - a device that scans the shape of your head with a laser and produces a 3D computer model. The first time I tried it the device went round and round my head but didn't produce any data. The person operating it said that the fuse on the scanning laser must have been blown and would need to be replaced. However, she soon realized that the problem was that the laser was simply turned off. We tried again but this time she started it while I was still climbing back into the chair, so I had to duck to avoid getting hit by the swinging arm. When it came out my head had no face. The third time we tried, it worked, but my glasses reflected the laser in such a way that the computer thought that my face had large spikes sticking out horizontally. I tried to save this image and email it to myself so I could put it up on this blog, but the computer's Internet connection didn't work. I then tried to take a picture of the computer screen showing the image with my cell phone, but it was out of batteries.

I did however get to talk to several professors, including Michael Heath, who knew two professors at Maryland that I have had classes with; Steven LaValle, who literally wrote the book on automatic algorithms for planning movements of robots, and Jeff Erickson, who told us all about what it's like to live in Illinois, including the fact that all employees of the State of Illinois (and this includes employees of UIUC) have to take mandatory "ethics training" every year. However, this "ethics training" must not be working very will, considering that apparently 3 out of the last 5 governors of Illinois have ended up in prison (and that's not even counting Rod Blagojevich). Also he said he knows someone who lives near the statehouse in Chicago, and he's actually seen politicians walk out of the statehouse carrying large bags full of cash.

After dinner Friday night, we had a scavenger hunt where we got into groups and had to find locations in downtown Champaign. It was problematic because a lot of the pictures we were supposed to take were of items located inside buildings, and whoever wrote it up forgot that we would be doing this at around 9:00 at night, when a lot of those buildings would be closed. But we were able to find lots of the buildings by putting them into a GPS.

Overall, I think that UIUC will be a place that I want to go for graduate school. I would probably choose it over University of Maryland just because it gives me a chance to go to a different palce, live in a different area, meet new people, etc. And despite being farther away from a major city than University of Maryland is, there still seems to be plenty to do, especially outdoor activities like bike trails and mountains. And that's not the only perk - the local public library even rents video games.

(P.S. I have no idea whether there actually is a Holiday Inn Express anywhere near UIUC. I just put that line in there because it was funny.)

Sunday, February 22, 2009

Weapons-Grade Bamboo

Apparently, there's an Amtgarder around here who has a "friend" that needs to get rid of a whole bunch of bamboo plants, so he's going to harvest them, dry them himself to make it "weapons-grade," and then sell it to other Amtgarders. The price? $15.00 per stick. Instead, I looked online for other options.

Cali Bamboo sells 10 foot long, 1 inch diameter poles for $2.25 each, but you have to buy 25 of them. I also don't know how much shipping costs. However, it appears you can also buy bamboo fencing and then separaate the poles, and according to my calcualtions that's actually a little cheaper than buying the poles separately. I don't know whether they're dried or not.

Bamboo Supply Company sells bales of 50 poles that are each 8 feet by 3/4 inch diameter for $48.50 per bale. That's about $1 per pole. Even if shipping is really expensive that still beats the $15 per pole. And this web site specifically says the bamboo is dried. I think a good strategy would be to buy a big bale, and then go to an Amtgard event and sell them. If people are really willing to pay $15 each, I see a very wide profit margin opportunity here.

Friday, February 20, 2009

Bad news, good news

I just heard back from Carnegie Mellon University, one of the places I applied for graduate school. I did not get in.

I did, however, receive a notice from the University of Maryland that I have been selected, on the basis of my academic record, as a "possible candidate for student speaker at the University Commencement Ceremony." I don't know how many other students got this letter or how many will be selected to speak. It said that if I'm interested I should send them a draft of my speech and it should be 2-4 minutes long.

It would be cool, the only thing is I'm not exactly sure what I would talk about. I do have until March 30 to send it in, so I will have some time to think about it.

Thursday, February 19, 2009

Textbook Economics

So, there's an interesting issue being discussed in the Diamondback.

Apparently, student activists are up in arms about the high price of textbooks and are working on trying to find ways to lower them. Some of these methods include:

- Mandating that professors allow older editions of textbooks or justify why they need the new editions.
- Post the ISBN numbers for textbooks so that students can comparison-shop online.
- Posting the book lists earlier so that people have more time to shop around.

There is only one slight problem. If students can comparison-shop, then more of them will buy their textbooks at places other than the University Book Center. And since profit from the University Book Center is used to run the Student Union and pay off the debt from its construction, if UBC profit decreases then student fees will have to go up in order to make up the money.

Amtgard and the SCA

"What do you expect? It's not like we have professional safety certifications or references to ASTM standards here."

A discussion on Electric Samurai, an Amtgard discussion forum, about how to convince a university that Amtgard is not a liability risk

"Rattan-cored Siloflex or Siloflex equivalent weapons shall be constructed using tubular materials meeting ASTM standard D-2239 or the international equivalent, with a pressure rating of 160 PSI or greater, having at least a 1 1⁄4 inch (31.8 mm) diameter on the outside and at least 1/8 inch (3.2 mm) walls, and having an inner core of rattan that fills the interior of the tubular material entirely."

- The SCA combat rules handbook


While going around to some bars on Route 1 trying to sign people up for Amtgard, a bartender told me that there was another group of people who "dress up like knights and fight" at a church a couple blocks away on Mondays. I went there and found that they were, in fact, an SCA group. SCA stands for Society for Creative Anachronism, and it's the main medieval reenactment organization with a focus on historical accuracy. So you won't find any magical spells or monsters there. Instead they use weapons - and these weapons are very different than the ones in Amtgard.

For one thing, they are much heavier. They are usually made of rattan, a kind of wood, with no padding whatsoever (except on the front tips of thrusting weapons like spears.) This of course means that the weapons hurt a lot, so they have to wear lots of armor. Minimum armor requirements include armor that covers almost of your body, including a helmet. (The requirements are very detailed - see the SCA combat handbook, pages 10 to 12.) Even still, people get hurt a lot. Several fighters proudly showed me their battle wounds. One person even said that they had once been hit by a particularly hard blow in the shoulder. A few days later nothing was apparently wrong with his shoulder, but there was a discoloration going down the length of his forearm. What actually happened was that he had bled internally, and later had to get the blood sucked out with a syringe.

But anway, when I asked to fight, I was told that I woul dhave to spend about two months just learning "footwork" beore they would even let me gear up and actually fight someone. Then it would be a couple more months before they would let me fight in an actual battle (rather than just practice.) And then if I want to do combat archery there's yet another process I have to go through.

But getting back to the original reason I was there, I told them about LARP and they told me in no uncertain terms that they "don't do LARP." However there was one person who was interested so I gave them the information about the web site.

Wednesday, February 11, 2009

Amtgard News #4

So, I started a new recruiting tactic: rather than giving out flyers, get people to give me their email addresses and sign up for a mailing list. So far I have 35 email addresses on my mailing list. Unfortunately, last time nobody who was on the mailing list came, although two people did say they tried to come but got lost. (I thought I was pretty clear where the park was because I posted up the link to the Google Map.) However I also discovered, when I put my own email address on the list (I set up a separate email address for the list) just to check whether the messages were getting sent out correctly, that the message was sent to me - but was directed to my spam folder. It is a good chance that this was caused by me writing "to unsubscribe from this mailing list, send mail to such-and-such address" at the bottom, as this is characteristic of a lot of spam emails. When I sent the same message to myself without that line, it got through.

I also came up with a new idea for my next Arts and Sciences project. It is a software program that is designed to be useful for foamsmiths (that's the term for people who create foam weapons like those used in Amtgard.) Features that I plan to include are:

- Create weapon designs: specify length, materials used, location of striking edges, etc. and automatically test new weapon designs for legality, weight, swingability, etc. before spending lots of time and materials building them.

- Set up production lists and automatically generate "shopping lists" of required materials - so you don't spend lots of money on materials you don't need or end up missing key components.

- Also: add your own material types, create shields and arrows in addition to weapons, compare different materials, and much more.

Friday, February 6, 2009

More on the Census Bureau

The problem I am working on for the Census Bureau is known as "unbiased controlled rounding." The problem is the following:


- A two-dimensional array of values, each in the range [0,1).
- The marginal counts - i.e. the sum of each row, and the sum of each column (and the grand total.) We know that the row sums and column sums are all integers, and they of course are equal to what the values in each column add up to.


- A randomized method of producing an array of zeros and ones the same size as the original array such that:

-- The expected value in each cell is the same as the corresponding value in the original array.

-- The row sums and column sums are the same as those in the original array (all the time, not just on the average.)


The main application of this problem is to reduce the risk of disclosure when releasing "microdata" - data with very small sample sizes (for example, all the businesses within a particular area that's ten blocks wide) Basically you want to round off the answers that the respondents gave so that an adversary cannot identify an individual respondent by their answers, but still preserve all the aggregate statistical properties (like mean, variance etc.) and avoid introducing bias.

As it turns out, the two-dimensional case has a known solution. The three-dimensional case (where in addition to row and column sums, you have sums going in the third direction, and you also have sums through planar "slices" of the cubical "array") does not, and in fact it is not always possible to solve.However, the goal is to get as close as possible to a solution (say, make sure that the marginal counts are off by at most 2) and do this in a reasonably computationally efficient way.


BONUS QUESTION: Solve the problem given above for the one-dimensional case (i.e., when the array just has one row in it, and the only marginal sum there is is the grand total). The solution to this problem gives an idea of some of the techniques used in higher dimensional solutions.