Re:The Cab Ride I'll Never Forget
by Retric on Thursday May 03, 2012 @09:53AM
(#3923830)
Occam's razor it's simply a way to compare two theory's with IDENTICAL results. AKA for all test data A and B produce the same results however A takes 10 days on a supercomputer to compute and B can be done by hand, I will go with B.
It also has a strong theoretical basis. Add enough epicycles and you can smooth out any irregularity's. But, over fit your data and you lose predictive capability.
PS: Preferring the more complex theory is unworkable, because E=mc^2+(10^(-900))m is going to give the same results for any test you could actually carry out. Having no preference has the same basic limitation.
It also has a strong theoretical basis. Add enough epicycles and you can smooth out any irregularity's. But, over fit your data and you lose predictive capability.
PS: Preferring the more complex theory is unworkable, because E=mc^2+(10^(-900))m is going to give the same results for any test you could actually carry out. Having no preference has the same basic limitation.