@FfKnighty @AndrewM_Webb @PeterZhizhin Why not use universal function approximators? 😉 They either learn „random mode mixing“ or an approximation of inverse transform sampling, depending on the architecture. https://t.co/6oGJ0VM1hI
18,691 followers
276 followers
RT @maier_ak: Can we use #deeplearning to produce random numbers of arbitrary distributions? Felix Horger's Distribution Remapping approxim…
18,691 followers
Can we use #deeplearning to produce random numbers of arbitrary distributions? Felix Horger's Distribution Remapping approximates known random number sampling strategies. Yet, it is able to do so without any explicit statistical model. https://t.co/vRX4xj