def decision_boundary(X, Y, w, b):
x_min, x_max = -2, 4
y_min, y_max = -5, 3
xx, yy = np.meshgrid(np.arange(x_min, x_max, .05),
np.arange(y_min, y_max, .05))
if len(w) > 2:
XX = polynomial_transform(np.vstack((xx.ravel(), yy.ravel())).T)
else:
XX = np.vstack((xx.ravel(), yy.ravel())).T
Z = np.dot(XX, w) + b
# Put the result into a color plot
Z = Z.reshape(xx.shape)
plt.contourf(xx, yy, Z > 0, cmap=plt.cm.Paired)
plt.axis('off')
# Plot also the training points
plt.scatter(X[:, 0], X[:, 1], c=Y, cmap='winter')
plt.axis([-2, 4, -5, 3])
plt.draw()
# Load the dataset
with np.load('./' + dataset + '.npz') as data:
X = data['X']
Y = data['Y']
algorithms = ['perceptron', 'logistic_regression', 'polynomial_regression']
algorithm = algorithms[1]
Xorig = X
if algorithm == 'perceptron':
eps = 1.
elif algorithm == 'logistic_regression':
eps = 5.
elif algorithm == 'polynomial_regression':
X = polynomial_transform(X)
eps = 100
nSamples = X.shape[0]
plt.ion()
w = np.zeros(X.shape[1])
b = 0.
for i in range(100):
if algorithm == 'perceptron':
# Predicted outputs
T = (np.dot(X, w) + b) > 0
C = T
else:
# Predicted outputs
# Load the dataset
with np.load('./' + dataset + '.npz') as data:
X = data['X']
Y = data['Y']
algorithms = ['perceptron', 'logistic_regression', 'polynomial_regression']
algorithm = algorithms[1]
Xorig = X
if algorithm == 'perceptron':
eps = 1.
elif algorithm == 'logistic_regression':
eps = 1. / np.amax(np.sum(X**2, 1))
elif algorithm == 'polynomial_regression':
X = polynomial_transform(X)
eps = 20 / np.amax(np.sum(X**2, 1))
nSamples = X.shape[0]
plt.ion()
w = np.zeros(X.shape[1])
b = 0.
T = np.zeros(nSamples)
C = np.zeros(nSamples)
sum_gradients = np.zeros(X.shape[1])
gradients = np.zeros(nSamples)
for i in range(100 * nSamples):
# Draw an example at random
