n_hidden_layers = 2 # number of hidden layer
# here 2 Hidden layer with 8 node each and 1 output layer with 3 node
#------------------------DEclaration of activation or Transfer function at each layer --------------------------------------#
# specify activation functions per layer eg: [ hidden_layer_1, hidden_layer_2, output_layer ]
activation_functions = [
symmetric_elliot_function,
] * n_hidden_layers + [sigmoid_function]
# initialize the neural network
network = NeuralNet(n_inputs, n_outputs, n_hiddens, n_hidden_layers,
activation_functions)
# network is Instance of class Neuralnet
# start training on test set one
network.backpropagation(training_one,
ERROR_LIMIT=.05,
learning_rate=0.2,
momentum_factor=0.2)
# save the trained network
network.save_to_file("trained_configuration.pkl")
# load a stored network configuration
# network = NeuralNet.load_from_file( "trained_configuration.pkl" )
# print out the result
for instance in training_one:
print instance.features, network.forwordProp(np.array(
[instance.features])), "\ttarget:", instance.targets
training_one.append(Instance(inp[i][0],inp[i][1])) #Encapsulation of a `input signal : output signal
#------------------------------------------------------------------------------
n_inputs = 4 # Number of input feature
n_outputs = 3 # Number of neuron output
n_hiddens = 8 # Number of neuron at each hidden layer
n_hidden_layers = 2 # number of hidden layer
# here 2 Hidden layer with 8 node each and 1 output layer with 3 node
#------------------------DEclaration of activation or Transfer function at each layer --------------------------------------#
# specify activation functions per layer eg: [ hidden_layer_1, hidden_layer_2, output_layer ]
activation_functions = [symmetric_elliot_function,]*n_hidden_layers + [ sigmoid_function ]
# initialize the neural network
network = NeuralNet(n_inputs, n_outputs, n_hiddens, n_hidden_layers, activation_functions)
# network is Instance of class Neuralnet
# start training on test set one
network.backpropagation(training_one, ERROR_LIMIT=.05, learning_rate=0.2, momentum_factor=0.2 )
# save the trained network
network.save_to_file( "trained_configuration.pkl" )
# load a stored network configuration
# network = NeuralNet.load_from_file( "trained_configuration.pkl" )
# print out the result
for instance in training_one:
print instance.features, network.forwordProp( np.array([instance.features]) ), "\ttarget:", instance.targets
