medium_spicy

4 ms·

1.
by medium_spicy 3y ago
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 represen
2.
by medium_spicy 3y ago
Haven’t been to Norway, then?
3.
by medium_spicy 3y ago
*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 sh
4.
by medium_spicy 3y ago
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 a
5.
by medium_spicy 5y ago
- This thread is about marketing. Did you do all of the marketing, or did an existing infrastructure perform tracking and serve ads for you? - What data support claims about your users’ experience? Conversions are not a good metric of user
6.
by medium_spicy 5y ago
I’m mostly a radar guy so I’m missing it — is it the fact that we only see the upper sideband, as though the signal is analytic? I don’t know this software (only a bit of GNUradio now and then) so I don’t know if the green bar up top is jus
7.
by medium_spicy 5y ago
This doesn’t apply to folks who want contacts for what’s in their profile, but: sometimes you’d like employers to know what relevant adjacent skills you have, without recruiters hitting them as search terms. For that purpose, I masked all o