Some useful math cheat sheets - take two ...
math-blog.com
math-blog.com
Your message comes off as if you had no idea what was wrong with what you were doing, even though OP pointed it out. And then you ask if it needs to be fixed. Are you trolling or something? Jesus.
==============
Edit: So it turns out that it's not RiderOfGiraffes' site, my fault for being temperamental, and misunderstanding. I just tend to get pissed off at people not giving credit.
I found the author's e-mail addresses if anyone wants to help set things right: acangiano at gmail dot com info at antoniocangiano dot com
> OP points out that you're not giving credit to the
> original authors. I think it's pretty obvious what
> you need to do. This is in the standard ethical
> code of conduct. Give credit to the original authors
> by maybe adding the author's name / website next to
> the link.
WTF? It's not my web page!It's not me who's not giving credit, it's not me who can change the page, it's not me who can do anything about it!
> Your message comes off as if you had no idea what was
> wrong with what you were doing, even though OP pointed
> it out. And then you ask if it needs to be fixed. Are
> you trolling or something? Jesus
It wasn't me doing anything, I'm asking if we, who don't control the page, can do something to redress the balance.Check your facts.
I have now submitted the link to the originals as an item in its own right: http://news.ycombinator.com/item?id=2346818
================================
Edit: OK - email sent to the site owner.
The covariance measures how "linearly related" two RVs are, not "how related". In the convolution formula for the sum of two RVs, you also need independence of the RVs for the convolution to be the valid approach. In the description about uncorrelated, there's a statement that "the reverse holds only for Gaussian RVs" which is not true, the word "only" needs to be taken out. (The sheet is trying to say the right thing on that one, it's just a language issue.) The N(0,1) is called a standard normal distribution, not just a normal distribution. The last equation on the sheet, for a vector normal distribution, forgets to invert the covariance matrix that appears in the exponential.