Executive Summaries

Posted February 4, 1999 at 1:55 am No Comments

Do you ever wonder how your classes manage to fill up an entire semester?  When I stop for a moment to take stock of the course material I’ve retained, it always seems like each of my courses could have been summed up in one 5-minute lecture.  Because I’m such a nice guy, I’m going to share with you the quick-and-dirty 5-minute version of each of the classes I’m taking right now.  That way you won’t have to waste any time taking them next semester.

  • 82-416, Contemporary Emerging French Literature
    – Swiss people are really depressed all the time.
    – It sucks to be old.
    – When you’re speaking French, don’t make grammatical errors.
  • 15-312, Programming Languages Design and Practice
    – ML is a really good programming language.
    – Lisp and Scheme really suck.
    – Java sucks too. Man, does it suck.
    – Don’t even get me started on how much C++ sucks.
  • 21-341, Linear Algebra
    – The term “n-tuple” is barbarous. Don’t ever use it.
    – The real numbers are a field. That lets you do a lot of crazy shit.
    – Linear Algebra textbook authors are complete morons.
    – You are bad at math.
  • 36-226, Probability and Statistics II
    – If you flip a coin ten times and get heads five times, a good bet would be that the probability of heads is one half.
    – But only a moron would use that intuitive method.  If you want to be a true statistician you have to calculate the first and second derivatives of the log likelihood function for the binomial distribution and solve for the unknown parameter by maximizing the score.  That way you can fill up a whole page with symbols, and still get 0.5 as your answer.  But now you get to call it “x bar”.
    – Actually that method is really crappy too.  What you really want to do is find an unbiased estimator and determine the sufficient statistic for the distribution.  Then you compute the expected value of the estimator given the sufficient statistic and that gives you a uniformly minimum variance unbiased estimator, which is of course far superior.  Also maybe instead of using the binomial distribution, you should use the normal approximation to the binomial distribution, so that you can look a bunch of shit up in those little tables in the back of the book, and also use a few more Greek letters, like sigma and mu.  But don’t worry, your answer will still be 0.5.

No Comments yet »

RSS feed for comments on this post. TrackBack URI

Leave a comment

XHTML: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

Powered by WordPress with Pool theme design by Borja Fernandez.
Entries and comments feeds.