8 Output(0.0), Delta(0.0), Net(0.0), cachedoutput(wantcached)
17 double SigmoidalNeuron::CalcOutput()
19 Net = Combiner->CalcOutput();
20 Output = Function->CalcOutput(Net);
27 Combiner->SetInput(input);
35 Function = boost::shared_ptr<GeneralActivationFunction>(
39 Function = boost::shared_ptr<GeneralActivationFunction>(
43 Combiner = boost::shared_ptr<GeneralLinearCombiner>(
virtual ~SigmoidalNeuron()
This activation function simply outputs its input.
virtual void SetInput(const GeneralLinearCombiner::tinvector &input)
Set the input neurons.
virtual void SetType(tneurontype type)
Set the type of neuron, determines the activation function.
SigmoidalNeuron(tneurontype type, bool wantcached=false)
Construct neuron with a known type.
The bipolar activation function is a common function in NN applications.
A linear combiner as a component of a neural network.
std::vector< boost::shared_ptr< GeneralNeuron > > tinvector
tneurontype
At the moment there are two types of neurons: bipolar and identity, they differ in their activation f...