2 ms·
You’re changing the subject! There is a really good problem in there. Why does this stuff work so well, for what classes of problems, and can we say anything a
by mturmon 2y ago
You’re changing the subject! There is a really good problem in there.
Why does this stuff work so well, for what classes of problems, and can we say anything about the fundamental energy/accuracy limits of this approach?
I think of it as an engineering problem, but there are mathematical rephrasings of this problem, as Michel Talagrand’s work has shown.
Don’t let the greed/stampede obscure the very interesting technical principles here.