Re:UC Irvine Machine Learning Repository
by sgt101 on Saturday November 22, 2014 @06:55AM
(#8645373)
I still think you are wrong.
There is no harm in developing a new algorithm, or coming up with new stories as to how it works, but to say that "data munging" will not advance the state of the art is to exclude the art! It's like saying that unless someone comes up with new physics it's impossible to build better spaceships.
And if Nasa were to build a better spaceship would that advance the state of the art in spaceship building?
Applying techniques to new problems involves data munging, and that is an important and useful task, but it is not difficult in the same way that doing new algorithmic work is, nor does it advance the state of the art.