module Ai4cr::NeuralNetwork::Cmn::MiniNetConcerns::TrainAndAdjust

Direct including types

Defined in:

ai4cr/neural_network/cmn/mini_net_concerns/train_and_adjust.cr

Instance Method Summary

Instance Method Detail

def calculate_error_distance #

[View source]
def derivative_propagation_function #

[View source]
def guesses_best #

[View source]
def init_net_re_train #

[View source]
def input_deltas #

[View source]
def input_deltas=(input_deltas) #

[View source]
def last_changes #

[View source]
def last_changes=(last_changes) #

[View source]
def load_outputs_deltas(outputs_deltas) #

[View source]
def load_outputs_expected(outputs_expected) #

[View source]
def output_deltas #

[View source]
def output_deltas=(output_deltas) #

[View source]
def output_errors #

[View source]
def output_errors=(output_errors) #

[View source]
def outputs_expected #

[View source]
def outputs_expected=(outputs_expected) #

[View source]
def set_deriv_scale_prelu(scale) #

Per Learning Style:


[View source]
def step_backpropagate #

[View source]
def step_calc_input_deltas #

Calculate deltas for hidden layers


[View source]
def step_calc_output_errors #

[View source]
def step_calculate_output_deltas #

Calculate deltas for output layer


[View source]
def step_load_chained_outputs_deltas(outputs_deltas) #

This would be a chained MiniNet's input_deltas e.g.: mini_net_A feeds is chained into mini_net_B So you would mini_net_A.step_load_chained_outputs_deltas(mini_net_B.input_deltas)


[View source]
def step_load_outputs(outputs_expected) #

[View source]
def step_update_weights(parallel = false) #

[View source]
def step_update_weights_v1 #

[View source]
def step_update_weights_v2 #

[View source]
def train(inputs_given, outputs_expected, until_min_avg_error = UNTIL_MIN_AVG_ERROR_DEFAULT) #

training steps

TODO utilize until_min_avg_error


[View source]