throwqwerty

>Nobody is forcing them to go to graduate school in California and accept this job. This isn't something that they were coerced into.

What is the significance of this point when literally no one is forced or coerced except for slaves? By this token absolutely no one should be striking for higher wages - they should all go work somewhere else. Oh wait...

>They do perform a service, sometimes teaching classes, sometimes grading papers, sometimes assisting with office hours. They get paid $2400/mo, putting them above median salary for full-time TAs.

Medían where? The entire United States? California? Santa Cruz? And exactly why is that relevant anyway? It's quite possible that the median wage isn't enough! And this would be part of the process for raising it.

>The fact of the matter is that research is still a demand driven field. This is why you'll see stipends and grants to assist students in higher demand fields, like computer science.

When you look at the average salary of a tenured UC professor (the kind that takes on grad students) and the average salary of appointed administrators (deans, provosts, etc) it's very hard to believe that the issue is lack of money and not allocation.

I also have no idea what cs/stem has to do with this. Most stem PhDs don't get more money either (most do not have fellowships nor their own grants).

I read a tweet on Reddit: if you think you made it through hard times okay but think it's okay for other people to go through the same hard times then you in fact did not make it through okay.

because they perform a service. they teach classes. that's literally what they were withholding - grades for classes they TA. so there is plenty of demand for that service. it has nothing to do with the labor market after they graduate. god some people just fall all over themselves to be smug.

lol about as cringy as corona beer being called corona beer (their power users have been called super hosts for years)

my mistake i didn't realize spectrogram and spectrum were distinct objects.

>Also, it’s computationally intractable to individually train 2^N weights.

that's a good point - i'd forgotten for a moment (because i'm so used to cooley-tukey fft) that in principle getting the spectrum involves a matmul against the entire vector. which brings up a potentially interested question: can you get a DNN to simulate the cooley-tukey fft (stride permutations and all).

why run a CNN over a spectrogram? besides what I said (convolutions in frequency domain are multiplications in time domain) an FT is linear. if classifying using those features were effective then your CNN would've learned the DFT matrix weights from the original signal.

but people aren't commodity goods so what exactly does that example prove?

a 1D CNN is an extremely simple RNN...

do people really run 2d convolutions on spectrograms? that seems rather backwards to me - why convolve in frequency space when you can just multiply in the time domain.

re regression tasks: sure I guess that's just an embedding basically right?

You wouldn't use a CNN for this because there's no spatial correlation to take advantage of. Maybe an RNN somehow but the basic problem is scoring (ie how melodic a song is) not classification so honestly I don't know you'd do this using neural networks.

I'd buy puts and write calls on all of the public unicorns in the bay because they're going to take a huge hit on productivity as soon as people start being hospitalized

i had a funny thought on reading bloomberg's summary of today: this is going to ruin trump because he'll handle this so poorly and he'll be up against a democractic primary candidate (either sanders or warren) that supports free healthcare

ah right right

I have a funny question. I have a FB internship this summer at the Menlo Park office. I know interns get health insurance but if I get Coronavirus while I'm doing the internship will they cover my treatment for the duration? I don't think my University provided health insurance will cover me while I'm in California...

imagine all the wind on the earth's surface right now. that's a vector field or gauge (as physicists call it). there are many units that it could denominated in (meters per sec, feet per second, etc). changing from one to the other is a gauge transformation (unit change, rescaling).

okay now suppose you want to compare the wind at two points. you can't just take the vectors at those two points and take a dot product because they're not in the same vector space (each of them is in a vector space tangent to the point they're anchored at). to figure out the issue with doing this watch this video

https://youtu.be/p1tfZD2Bm0w?t=170

you have to perform parallel transport. one way to do this is to track how much moving the vector changes it (messes with its orientation). that's the connection and the covariant derivative. you can then take a look at how that connection itself changes over the surface of the earth. if it's curl free (recall curl free vector field from multivariable calculus) then you don't have eddies in your wind charts i.e. you can't get faster by going around in a loop.

absolutely none of this applies to currency or equities because the surface of the earth is smooth and curved. neither of this is true about equities or securities or currency.