Yes, it is absolutely mind blowing that this taking place. One of the best ways of the computational waste (which will eventually bring down the hype) is the attempt to implement simple models that conventional stats can fit effortlessly as Deep Learning algorithms. How can one trust the tool, if it leads to overcomplicated , wasteful compute solutions?
Yes, it is absolutely mind blowing that this taking place. One of the best ways of the computational waste (which will eventually bring down the hype) is the attempt to implement simple models that conventional stats can fit effortlessly as Deep Learning algorithms. How can one trust the tool, if it leads to overcomplicated , wasteful compute solutions?
(I briefly spoke about this at mstdn https://mstdn.science/@ChristosArgyrop/110702155833967223)