5.01.2012

Day 344

One final down, one to go.


This morning started pretty normally.  I made a salad for lunch, but I didn't have any eggs or chicken to put on it... so I just used leftover stir-fry (minus the rice).  It was ok.


This morning in the computer lab I studied with Timbo and Matt.  Mostly going over equations and explaining some of the questions from past quizzes to Matt.  Yeah, we studied for a couple hours then went to go take it.


The test went ok.  There were two questions that I didn't know how to do, but I gave them my best shot.  One of the questions was poorly worded, so I gave it my best shot.  I didn't want to go up and ask him because one student had already been up a few times and Dr. Foley was starting to get annoyed.  I did go up on another question, but only because the question literally did not even ask a question.  So I gave it my best shot.  We'll see how it goes!  I finished in about an hour and just couldn't work the other problems any differently... so I put an answer and was done.


I went up to the computer lab after the test and just let my brain relax for a while.  I also got signatures from the necessary people for my grading and lab hours.  Then I had lunch and went back to the computer lab.  I did some studying, then discussed probability with David.  We talked about some of the fundamental problems with actuarial science, whose problems actually lie in the problems of statistics: how do you make your model fit the data?  For instance, in my field of study we constantly deal with the probabilities of surviving or dying over a specific time period.  These are measured by insurance companies at the end of every year.  But often we need to interpolate what happens and figure out what happens during the year.  So we make theorems and symbols and theoretical measurements... measurements that make our theories work perfectly.  But there is no such thing as a hazard rate (probability that something will break/die at a particular instance), but we use them constantly in our calculations.  So sometimes we have weird tautologies that are actually counter-productive.  Sometimes we have these circularly reasoned theorems... "We need A to go to C, but we don't know how it happens.  So let's make B such that A goes to B and B goes to C.  Now A goes to C!"  I don't want to make it sound like the math isn't valid.  These are common procedures in modeling.  The rules of math are bound by the syntax of logic and cannot be wrong.  The trouble is figuring out how the real data fits in our mathematical models.  If we could measure probabilities on every person at every single point in time, then we could create these (currently) fictional instantaneous probabilities.  But until that happens (not in my lifetime, anyhow), then we rely on theory.  And so far theory has not led us astray, provided we follow it.




Oh, and it rained today.




Anyhow.  When I came home I did some packing, dish washing, then did a little probability.  Packing was hard because most of the stuff I need I still use regularly.  So I'm guessing most of my packing will be tomorrow after I get off work.


So yeah.  I wasn't feeling to great so I didn't do much after dinner.  I'm just going to studying for a couple hours tomorrow in the computer lab, then go for it.  It's bed time now.  So good night everyone.

No comments:

Post a Comment