a test on genomes, more specifically genomes defining connections in a neural network notes: neural output: tanh(sum(inputs)) connection weights -4.0 - +4.0 (define that in genome) inputs produce 0 - 1.0 action neuron calculation: tanh(sum(inputs)) (which will come out to be -1.0 to + 1.0) (this is because the result of this calculation is the chance of the neuron firing) output = output * connection weight
Took some massive inspiration from this report on neural networks and natural selection: https://www.youtube.com/watch?v=N3tRFayqVtk I also took some notes around 23:00 currently this is nothing but some notes
