class SHAInet::Neuron
- SHAInet::Neuron
- Reference
- Object
Defined in:
shainet/basic/neuron.crConstructors
Instance Method Summary
-
#activate(activation_function : ActivationFunction = SHAInet.sigmoid) : Float64
This is the forward propogation Allows the neuron to absorb the activation from its' own input neurons through the synapses Then, it sums the information and an activation function is applied to normalize the data
- #activation : Float64
- #activation=(activation : Float64)
- #bias : Float64
- #bias=(bias : Float64)
- #clone
- #gradient : Float64
- #gradient=(gradient : Float64)
- #gradient_batch : Float64
- #gradient_batch=(gradient_batch : Float64)
- #gradient_sum : Float64
- #gradient_sum=(gradient_sum : Float64)
-
#hidden_error_prop : Float64
This is the backward propogation of the hidden layers Allows the neuron to absorb the error from its' own target neurons through the synapses Then, it sums the information and a derivative of the activation function is applied to normalize the data
- #id : String
- #id=(id : String)
- #input_sum : Float64
- #input_sum=(input_sum : Float64)
-
#inspect
Returns an unambiguous and information-rich string representation of this object, typically intended for developers.
- #m_current : Float64
- #m_current=(m_current : Float64)
- #m_prev : Float64
- #m_prev=(m_prev : Float64)
- #n_type : String
- #n_type=(n_type : String)
- #prev_bias : Float64
- #prev_bias=(prev_bias : Float64)
- #prev_delta : Float64
- #prev_delta=(prev_delta : Float64)
- #prev_delta_b : Float64
- #prev_delta_b=(prev_delta_b : Float64)
- #prev_gradient : Float64
- #prev_gradient=(prev_gradient : Float64)
- #randomize_bias
- #sigma_prime : Float64
- #sigma_prime=(sigma_prime : Float64)
- #synapses_in : Array(SHAInet::Synapse)
- #synapses_in=(synapses_in : Array(SHAInet::Synapse))
- #synapses_out : Array(SHAInet::Synapse)
- #synapses_out=(synapses_out : Array(SHAInet::Synapse))
- #update_bias(value : Float64)
- #v_current : Float64
- #v_current=(v_current : Float64)
- #v_prev : Float64
- #v_prev=(v_prev : Float64)
Constructor Detail
Instance Method Detail
This is the forward propogation Allows the neuron to absorb the activation from its' own input neurons through the synapses Then, it sums the information and an activation function is applied to normalize the data
Returns an unambiguous and information-rich string representation of this object, typically intended for developers.
This method should usually not be overridden. It delegates to
#inspect(IO) which can be overridden for custom implementations.
Also see #to_s.