10/12/16

1. 10/12/16
1. Slept in a bit, went to StatMech. More engines. Gaah! When will this review be over?
2. First physics Math Club meeting in SPS lounge. Brandon, Jacob W. me, new guy named Keegan who’s Jacob’s friend, and Tristan from 33x. We tried to prove that any vector field with zero divergence could be written as a curl, no luck. Then Jacob talked about his current math research, which was really fascinating. One problem I’d like to think about more.
3. The setup is, given a tournament--a graph whose vertices are people and whose directed edges represent who beats whom in 1v1 matchups--you can construct many different sets S of voting ballots Bi, each Bi has a ranking of the people x1, … xN, such that xi beats xj in the tournament, if and only if the number of ballots where xi > xj exceeds the number where xi < xj.
4. When you put all the candidates together though, there is a different, overall criterion for winning; candidate xi beats candidate xj iff xi is ranked first on more ballots than xj.
5. The question is, can you construct a set of ballots for any tournament of 1-1 results, where, if you look at only the first element of each ballot’s ranking (that voter’s top candidate, so to speak), you can make any person win the vote, overall?
6. This is like asking, can you put enough information in the 2nd through nth entries of each ballot’s ranking, to wipe out the entries in the 1st column that determine the overall winner, when you’re considering 1v1 matches.
7. Anyway, enough math! After that was the SPS meeting, which was a bit absentminded (partly me). Devin is stepping up in his role as Halloween Party captain though, which is great.
8. Stayed in the SPS lounge, working on E&M homework, which is still annoying review. Even more annoying was the fact that it’s hard to set up Mathematica’s system for doing units calculations right. So I spent like an hour cursing my ancestors and looking through Mathematica help / support / random StackOverflow crud.
9. Finally went outside, which improved my mood considerably. There however, I was confronted by a super sock zombie. I thought I was through, done, sunk! But I hid behind my table and fended him off with my own socks. Eventually he got careless and ran out of socks (he had a lab in 5 minutes), and I rushed him and got the stun.
10. Unfortunately, my glorious human self didn’t make the Wednesday HvZ mission. I hustled to CSE building and did some quick physics tutoring for a friend from Lavin. Then I met Andrew’s research professor, Prof. Suciu. I had done just enough prep to know what was up, but not overthink things so much that my natural math curiosity couldn’t express itself. I basically grilled the guy on what it meant to factor databases as matrices. He seemed happy with my enthusiasm, invited me to work on the project after first doing some databases homework and then meeting with Andrew. I really liked him. Friendly, explained the context of our problem within his own research and within CS. And I liked what I might be working on. The intersection of an untapped obscure region of math (low-rank tensor factorization) with machine learning! :D :D
11. So that was exciting! Afterwards I went to the advanced swing lesson and really enjoyed myself. Skipped out on techniques because if I schedule all my time I don’t pay attention to the life I’m living :/
12. Back to 2104, board game night! Got food at the 8 and ate it while we played card games, some Set (crushed Evan, yeah!) and Bananagrams.
13. Went out to the IMA, ended up just playing 1v1 at Denny field, which saved me 20 minutes of walking. Got wrecked by this energetic Asian dude Luke, who invited me to play on his team next quarter. Very positive guy, would like to play with him more.
14. Returned to 2104 to finish E&M homework, book buses home for a day between classes and DubHacks. Got into an interesting conversation with some floormates, one of whom had to deal with a good friend who wanted to be a boyfriend but it wasn’t going to work out.
15. Finished the E&M hw, wrote this log and went to bed.