class MicroGrad::Neuron(T)

• MicroGrad::Neuron(T)
• Reference
• Object

Overview

Base neuron, always linear; use the sub-classes to enable non-linear activation using the desired function.

Included Modules

• MicroGrad::Common(T)

Direct Known Subclasses

• MicroGrad::Neuron::ReLU(T)
• MicroGrad::Neuron::Sigmoid(T)
• MicroGrad::Neuron::TanH(T)

Defined in:

micrograd/nn.cr

Constructors

• .new(num_inputs)

  Create a linear neuron with num_inputs

Instance Method Summary

• #activate!(inputs) : Value(T)

  Activate the neuron, layer or network with the given inputs

• #b : Value(T)
• #parameters : Array(Value(T))

  Retrieve references to all the parameters as an array

• #to_s(io)

  Prints a concise string representation, typically intended for users, to io

• #w : Array(Value(T))

Constructor Detail

Instance Method Detail