So I’m working with propensity scores again and just thought I’d revisit the topic again. Also thought maybe I’d write another poem again, so here goes. Clears throat …
They could be low.
So that your answer is no.
But they could be high.
So you have a yes, oh my!
Propensity scores they are.
And they can tell Randy how far.
Anton wondered, yes I fit them in.
This example. Why? Is that a sin?
But anyway, till next time.
Maybe I’ll give something else with rhyme.
Maybe you could indeed fair better.
But we’ll see next week what’s the matter.
Yeah, will work on that too … but anyway, till next time!
So as promised, we will give an example of available case analysis or AC, as I attended such a talk during my Seattle conference last week. And what this means is we don’t throw out entire entries but keep any entries with at least some information that could help us estimate the parameters. Kind of like Randy may not know the specific city that Anton lives in in a certain dimension, but he may know Anton’s occupation and marital status and can determine if his target is in a good or a bad dimension based on that. Kinda like that. Well, anyway that’s my exciting news for today. That and my dissertation advisor friended me on facebook. Well, that’s exciting to me at least! I just wonder if he also noticed that I’m friends with someone who was on Game of Thrones (thus far!) and with someone who worked with Drew Barrymore (again, thus far!). Well, he might find out eventually and then he might ask me how did that happen. To which, I may reply …
That’s okay. He’s used to that face. That’s the face I had for 85% of graduate school. But anyway … till next time!
Hey there! So just came back from my stats conferences and have the perfect topic for next week — but until then, I just noticed that there’s a movie theater next to the Washington State Convention Center.
So, you know, if the APS March meeting is ever held in Seattle and any of the physicists want to wonder to see a syfy movie that could be playing there based on the multiverse theory, then … yeah, maybe still my jetlag talking. But ya never know!
Hi! So I was going to skip a week since I am going to Seattle for the Joint Statistical Meetings, the same conference that was held in Boston last year, but then I figured why would I do that if this is the week Matt Damon could visit my blog! Matt Damon or John Krasinski. Speaking of Krasinski, let us continue the example from last week and talk about how we determine the points for each one of our factors. Well, we can do that by looking at the rank of our regression parameters. Now, say, the regression parameter corresponding to the actor cast as Randy is a 10. Then it is 10 in our model if the actor is Krasinski and 0 if it is someone else, as we are multiplying the parameter to a binary variable. Now what the regression parameters for the Jane and Anton roles? Well, they most probably are 5 and 3, respectively. As 5/10 = 50% or 50 points and 3/10 = 30% or 30 points. And what else could these oh-so-cool-oh-shush-yes-they-are regression parameters tell us? Well, that’s for next time! Until then, here’s to hoping that by just some strange chance of luck I do run into Krasinski in Seattle and talk up my book with him.
Oh shush … again, it could happen! Or at least it could happen with another meeting attendee from Brown, his alma mater. Just sayin’!
Out of This Dimension 