whilst this might look simple from the outside its slightly more complicated on the inside. to be clear: the ai knows what you picked, its just learning how to react to it
This uses a simple neuronal network with no hidden layers whatsoever, 3 input neurons beeing the pick you made (the first neuron stands for rock, the second for paper, and the 3rd for sissors.) it is beeing trained in 30,000 epochs based on a training set consiting of 3 input and 3 ouputs (the pick the "enemy" made as input and the pick that would win against it as output). this took me around 30 minutes to make because i lately started doing ai stuff and felt like implementing one in scratch as a little challenge but it was way easier than expected. This simply uses the delta rule, and doesnt use multi layer backwards propagation.
