>>13720208>>muh subjective probability updates from prior assumptionsYea, believe it or not all pattern recognition like in the human brain is due to Bayesian inference. We imagine one factor is removed or added and simulate the outcome given that new scenario. IT is subjective and it is imperfect, but neutral networks are just functions. They look spoopy, but you could always just get the same effect with a matrix, but matrix equations are spoopky enough to get published in Nature, right?
