This is a simple 2-2-1 neural network to solve the XOR problem. You can custumize the activations, number of neurons in the hidden layer etc. This program will use multiple CPU cores to speed up learning.
eval [input 1] [input 2] -> uses the NN to predict the Xor using current wieghts. Takes 2 integers
btrain [epochs] -> trains for given amount of epochs. Takes 1 integer.
end -> terminates the program. Takes no inputs.
print [variable] -> prints given variable. Takes 1 string (dont use quotes).
Variables
- x
- y
- w1
- b1
- w2
- b2
- samples