A neural network written in go
It is different from a traditional backprop network as it tryes to adjust the weights in a more "organic" way.
The activation functions and cost functions are rather trad, but the weights in the net are updated by random, directed movements.
- It would be interesting to take a look at golang. This is my first golang testout
- I am curious if I can combine an algorihm which combines natural selection programming and neural networks
- Coming up some time