n_samples = 400
sigma = 2e-1
X, T = datasets.make_moons(n_samples, noise=sigma)
half = int(n_samples / 2)
X_train = X[0:half, :]
T_train = T[0:half]
X_test = X[half:, :]
T_test = T[half:]
# Fit the neural network
max_iter = int(1e3)
tol = 1e-5
nn = NeuralNetwork(
[Layer('Tanh', (2, 5)),
Layer('Tanh', (5, 1))], max_iter=max_iter, tol=tol)
nn.fit(X_train, T_train)
# Predcition
prediction = nn.predict(X_test)
score = nn.score(X_test, T_test)
print(score)
# Visualization
plt.scatter(X_train[:, 0],
X_train[:, 1],
s=40,
zorder=10,
c=T_train,
import numpy as np
from matplotlib import pyplot as plt
from sklearn import datasets
from mlpy.model import Layer, NeuralNetwork
# Generate dataset
n_samples = 400
sigma = 2e-1
X, T = datasets.make_moons(n_samples, noise=sigma)
# Fit the neural network
max_iter = int(1e2)
tol = 1e-5
nn = NeuralNetwork(
[Layer('Sigmoid', (2, 3)),
Layer('Sigmoid', (3, 4)),
Layer('Sigmoid', (4, 5)),
Layer('Sigmoid', (5, 1))
], max_iter=max_iter, tol=tol)
nn.check_gradient(X, T)
