2012-04-25
There isn't really a lot to put here today.
Big thing: Continuing with the NaiveBayes.py assignment from the NLP class. Adding in an actual NB algorithm. (It's in the Private folder
NaiveBayes.py).
Another big thing: coding project with Sam and Isaac for Graph Theory/Four-Color Theorem. Team programming with version control through GitHub.
Also. Today's the deadline for Final Plan App, if I recall correctly. I totally forgot until I was looking through the calendar. I haven't been on my game with these big deadlines... I submitted it to you this morning.
From what I remember from your comments on my last submission:
- The "analysis" of the translation software:
- It is being done by "metrics" – yes, that's a fancy buzz-word – but both of the ones I'm using compare the machine and human in a fancy precision/recall system. I could type out a bunch of formulae, but for one, they're ones we've seen already in Tutorial a bajillion and one times, and also, what's the relevance here?
The two that I'm using are called BLEU (BiLingual Evaluation Understudy) and Meteor (Metric for Evaluation of Translation with Explicit ORdering). (
wikipedia: BLEU and
wikipedia: METEOR,
http://cs.cmu.edu/~alavie/METEOR/) Then there's the one that I made up, that isn't really the best, but it gives an idea, and also (interestingly) supports the rest of my data.
Also stat. tests are being run in R (partly because of the Stats class and Stats Workshop).
- I'm bringing in a bunch of translation theory. From both pre- and post-machine translation days.
- Plan Summary:
- Now one line: "An investigation of natural language processing and machine translation."
- As far as courses/tutorials for next year:
- Statistics Workshop (which I've already talked to Matt about)
- Sociophonology
- Probably something AI/Neural Net-related; paper-tracking tutorial.
- Do/can my exams count as a tutorial?
I need a total of 17 (9 Plan Credits) credits to reach the 120 necessary to graduate. I'm guessing about 1cr/10% if I did tutorials in Plan-paper-y things.
Next assignment (for the Tutorial – not Coursera): Final write-up.
Next Coursera assignment: CYK (Cocke-Younger-Kasami) Parsing. This will be the last one I do while we have classes.