return accuracy
if __name__ == "__main__":
# set random number generator seed
np.random.seed(NUMERO_DE_MATRICULA)
# set floating point formatting when printing
np.set_printoptions(formatter={"float": "{: 0.6f}".format})
# load data
x = DataSets.NOME_DO_DATASET.input
d = DataSets.NOME_DO_DATASET.output
# define the network parameters
n = TAXA_DE_APRENDIZADO
g = ActivationFunctions.FUNCAO_DE_ATIVACAO
# create the neural network
nn = Perceptron(n, g)
# train the neural network
w = nn.train(x, d)
# evaluate the neural network
acc = nn.evaluate(w, x, d)
# plot epoch versus error data
PlotUtils.plot(nn.plot_data_x, "epoch", nn.plot_data_y, "error")
correct = correct + 1
accuracy = float(correct) / float(total)
print('Accuracy: {:.2f}% ({}/{})'.format(100.0 * accuracy, correct,
total))
return accuracy
if __name__ == '__main__':
# set random number generator seed
np.random.seed(42)
# set floating point formatting when printing
np.set_printoptions(formatter={'float': '{: 0.6f}'.format})
# load data
x = DataSets.BLOOD_TRANSFUSION.input
d = DataSets.BLOOD_TRANSFUSION.output
# create the neural network
nn = Adaline()
# train the neural network
w = nn.train(x, d)
# evaluate the neural network
acc = nn.evaluate(w, x, d)
# plot epoch versus error data
PlotUtils.plot(nn.plot_data_x, 'epoch', nn.plot_data_y, 'mse')
for i in range(0, len(x)):
y = self.test(w, x[i])
if (y == d[i]):
correct = correct + 1
accuracy = 100.0 * (float(correct) / float(total))
print(f"Accuracy: {accuracy:.2f}% ({correct}/{total})")
return accuracy
if __name__ == "__main__":
# load data
x = DataSets.TIC_TAC_TOE_ENDGAME.input
d = DataSets.TIC_TAC_TOE_ENDGAME.output
# define the network parameters
n = 1e-4
g = ActivationFunctions.heaviside
e = 1e-10
# create the neural network
nn = Adaline(n, g, e)
# train the neural network
w = nn.train(x, d)
# evaluate the neural network
acc = nn.evaluate(w, x, d)
# plot epoch versus error data
PlotUtils.plot(nn.plot_data_x, "epoch", nn.plot_data_y, "mse")
y = self.test(w, x[i])
if (y == d[i]):
correct = correct + 1
accuracy = float(correct) / float(total)
print('Accuracy: {:.2f}% ({}/{})'.format(100.0 * accuracy, correct, total))
return accuracy
if __name__ == '__main__':
# set random number generator seed
np.random.seed(42)
# set floating point formatting when printing
np.set_printoptions(formatter={'float': '{: 0.6f}'.format})
# load data
x = DataSets.LOGIC_GATE_AND.input
d = DataSets.LOGIC_GATE_AND.output
# create the neural network
nn = Perceptron()
# train the neural network
w = nn.train(x, d)
# evaluate the neural network
acc = nn.evaluate(w, x, d)
# plot epoch versus error data
PlotUtils.plot(nn.plot_data_x, 'epoch', nn.plot_data_y, 'error')