Welcome to the Gallery – a place where you'll find hope and strength through the healing power of art and the universal reach of technology –
a place where you can
Connect, Create, and Thrive!
Interested in partnering with us?
Click here to find out how!
Displaying (2) Comments | Comment on this piece | Report objectionable art
Did I get involved? Don't reebmmer that. Or maybe you mean dhogazaAnyway on McI doubling up yes, this was discussed in passing in a couple of my posts. Mann's PCA algorithm involved short-centering *and* scaling (corresponds to correlation matrix). M&M performed conventional PCA with covariance matrix i.e. series length centering but also without scaling, a second difference from what Mann did. Huybers compared both centering conventions, but kept Mann's scaling. This showed that the bias in PC1 from short-centering alone was significantly less than that reported by M&M with their double change. In fact I think *each* of these two differences in Mann's PCA produces about the same amount of bias. M&M even mention in passing that some bias in PC1 came from the scaling choice (nevertheless a defensible choice for Mann), but never quantified or showed it. Wonder why? They left that to Huybers.
By: | Dec 04, 2012 | Report Comment
You are the ultimate son! Spider spray ineedd! Seriously, I hope you have a wonderful visit with your mom! It's wonderful when grown children and their moms can visit, catch up, visit with the grandchildren and just have a great time.Now, you're really not going to make your mom sleep in that tent, are you? (MD) Only if she tries to make me eat that chicken soup of hers.
By: | Jul 21, 2012 | Report Comment
abstract
Wacky kid 7
It draws you in.
There are 16 pieces of art in this thread