4 ms·

I am the author of this tutorial. If you are interesting in contributing a section to this tutorial, please get in touch. Some suggested topics: survival analys

I am the author of this tutorial. If you are interesting in contributing a section to this tutorial, please get in touch. Some suggested topics: survival analysis, mixture models, classification, time series models... Twitter @markdregan

gedrap 11y ago
Just wanted to say thanks a lot for taking the time to write it! :)
- markdregan 11y ago
  My pleasure. I learned a lot in writing it.
nl 11y ago
Are you @markdregan or @thinkvein? The @markdregan seems to indicate @thinkvein is you now?
- markdregan 11y ago
  Both are me. @markdregan is my choice handle. @thinkvein was for a blog I used to write.
kriro 11y ago
Looks pretty nice, good job. I think survival analysis is a very underrated tool. I'm working in UX now and there's a lot of test setups were survival analysis makes a lot of sense but isn't used (mothly because people don't know it).
- lookACamel 11y ago
  Would you like to expand on that?
ryanmonroe 11y ago
A few pedantic notes It would be nice if random variables were denoted by capital letters to distinguish them from particular observations. In the section about regression you should have the conditional mean of Y equal to \beta X, rather than the overall mean. Of course, this doesn't really matter too much since the substance of the tutorial is correct. I thought I'd mention it though because sloppy notation can undermine your credibility in some circles, and the equations are the most notable object when you're just skimming the page. More (very) pedantic notes: Usually, if you're going to express a sum with \cdots you put A+B+\cdots+Z (with plus signs on both sides of the ellipse for clarity) The probability operator is usually denoted with a capital P The log function shouldn't be italicized, you can just type \log to avoid this Distribution names in statements like y~Pois(\mu) aren't usually italicized.

gedrap 11y ago

Just wanted to say thanks a lot for taking the time to write it! :)

markdregan 11y ago
My pleasure. I learned a lot in writing it.

nl 11y ago

Are you @markdregan or @thinkvein? The @markdregan seems to indicate @thinkvein is you now?

markdregan 11y ago
Both are me. @markdregan is my choice handle. @thinkvein was for a blog I used to write.

kriro 11y ago

Looks pretty nice, good job. I think survival analysis is a very underrated tool. I'm working in UX now and there's a lot of test setups were survival analysis makes a lot of sense but isn't used (mothly because people don't know it).

lookACamel 11y ago
Would you like to expand on that?

ryanmonroe 11y ago

A few pedantic notes It would be nice if random variables were denoted by capital letters to distinguish them from particular observations. In the section about regression you should have the conditional mean of Y equal to \beta X, rather than the overall mean. Of course, this doesn't really matter too much since the substance of the tutorial is correct. I thought I'd mention it though because sloppy notation can undermine your credibility in some circles, and the equations are the most notable object when you're just skimming the page. More (very) pedantic notes: Usually, if you're going to express a sum with \cdots you put A+B+\cdots+Z (with plus signs on both sides of the ellipse for clarity) The probability operator is usually denoted with a capital P The log function shouldn't be italicized, you can just type \log to avoid this Distribution names in statements like y~Pois(\mu) aren't usually italicized.