Neural network primitives
The neural network primitives demonstrate logic functions using classical artificial neurons and McCulloch-Pitts neuron.
- This is a McCulloch-Pitts neuron. The activation of this neuron is binary. That is, at any time-step, the neuron either fires, or does not fire.
- This neuron will output the sum of the weighted inputs, as a floating value.
- Binary threshold of the input.
- Compute the Sigmoid function, defined as
where r is the learning_rate.