The similarity of Wolfram graphs made me think of neural nets. Is there anything in the Wolfram physics system similar to back propagation? In nature, all complex systems have some form of feedback or back propagation.
Whether dynamic or static, discrete or continuous, a neural network is a probability density function. Back-propagation is a means of determining weights in a static feed-forward neural network. GE employed them in clothes dryers for automatic shut-off back in the 90’s, essentially replacing Kalman filters.
So, how to use back propagation in Mathematica?
See the free neural net courses offered by Wolfram. Alternately, pick up a textbook and learn about probability density functions.
May you send me some links? Thanks.
Okay. Thanks a lot!