Short answer: if the inputs can be represented well on the Fourier basis, yes. I have a patent in process on this, fingers crossed.
Longer answer: deep learning models are usually trying to find the best nonlinear basis in which to represent inputs; if the inputs are well-represented (read that as: can be sparsely represented) in some basis known a-priori, it usually helps to just put them in that basis, e.g., by FFT’ing RF signals.
The challenge is that the overall-optimal basis might not be the same as those of any local minima, so you’ve got to do some tricks to nudge the network closer.
*can manage on its own well. The market manages - explicitly or through regulatory capture and toadies - a hell of a lot of things that the state probably should, like the Texas power grid. Markets are rarely efficient except in the very short term and very small scale.
No, this is only true if that hyperplane contains the origin; imagine an infinite number of hyperplanes that contain A and B; there are an infinite number of such planes for dimensions higher than 2. Now imagine the same, but connecting O and C; most of those AB hyperplanes are not orthogonal to those OC hyperplanes, it’s only coincidence if they are, though for dimensions higher than 2 you can always find a point C that happens to lie along the orthogonal line from the AB plane to the origin.
Longer answer: deep learning models are usually trying to find the best nonlinear basis in which to represent inputs; if the inputs are well-represented (read that as: can be sparsely represented) in some basis known a-priori, it usually helps to just put them in that basis, e.g., by FFT’ing RF signals.
The challenge is that the overall-optimal basis might not be the same as those of any local minima, so you’ve got to do some tricks to nudge the network closer.