iter1
This commit is contained in:
DmitriyAntonov
parent
94a76f47d8
commit
9bf1c4845a
@ -1,59 +1,97 @@
|
||||
import numpy as np
|
||||
from matplotlib import pyplot as plt
|
||||
from matplotlib.colors import ListedColormap
|
||||
from sklearn.model_selection import train_test_split
|
||||
from sklearn.preprocessing import StandardScaler
|
||||
from skimage.metrics import mean_squared_error
|
||||
from sklearn.datasets import make_moons, make_circles, make_classification
|
||||
from sklearn.neural_network import MLPClassifier
|
||||
from sklearn.linear_model import LinearRegression, Ridge
|
||||
from sklearn.model_selection import train_test_split
|
||||
from sklearn.pipeline import make_pipeline
|
||||
from sklearn.preprocessing import StandardScaler, PolynomialFeatures
|
||||
|
||||
X, y = make_classification(
|
||||
n_features=2,
|
||||
n_redundant=0,
|
||||
n_informative=2,
|
||||
random_state=0,
|
||||
n_clusters_per_class=1
|
||||
)
|
||||
|
||||
X, y = make_classification(n_features=2, n_redundant=0, n_informative=2, random_state=0, n_clusters_per_class=1)
|
||||
rng = np.random.RandomState(2)
|
||||
X += 2 * rng.uniform(size=X.shape)
|
||||
linearly_dataset = (X, y)
|
||||
moon_dataset = make_moons(noise=0.3, random_state=0)
|
||||
circles_dataset = make_circles(noise=0.2, factor=0.5, random_state=1)
|
||||
|
||||
datasets = [moon_dataset, circles_dataset, linearly_dataset]
|
||||
|
||||
"""
|
||||
Данные:
|
||||
· moon_dataset
|
||||
· circles_dataset
|
||||
· linearly_dataset
|
||||
"""
|
||||
for ds_cnt, ds in enumerate(datasets):
|
||||
X, y = ds
|
||||
X = StandardScaler().fit_transform(X)
|
||||
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=.4, random_state=42)
|
||||
X_train, X_test, y_train, y_test = train_test_split(
|
||||
X, y, test_size=.4, random_state=42
|
||||
)
|
||||
"""
|
||||
Модели:
|
||||
· Линейную регрессию
|
||||
· Полиномиальную регрессию (со степенью 3)
|
||||
· Гребневую полиномиальную регрессию (со степенью 3, alpha = 1.0)
|
||||
"""
|
||||
|
||||
alphas = np.logspace(-5, 3, 5)
|
||||
# Линейная
|
||||
linear_regression = LinearRegression()
|
||||
linear_regression.fit(X_train, y_train)
|
||||
linear_predictions = linear_regression.predict(X_test)
|
||||
linear_mse = mean_squared_error(y_test, linear_predictions)
|
||||
|
||||
current_subplot = plt.subplot(3, 5 + 1, i)
|
||||
# Полиномиальная (degree=3)
|
||||
poly_regression = make_pipeline(PolynomialFeatures(degree=3), LinearRegression())
|
||||
poly_regression.fit(X_train, y_train)
|
||||
poly_predictions = poly_regression.predict(X_test)
|
||||
poly_mse = mean_squared_error(y_test, poly_predictions)
|
||||
|
||||
current_subplot.scatter(X_train[:, 0], X_train[:, 1], c=y_train, cmap=cm_bright)
|
||||
current_subplot.scatter(X_test[:, 0], X_test[:, 1], c=y_test, cmap=cm_bright, alpha=0.6)
|
||||
# Гребневая (degree=3, alpha=1.0)
|
||||
poly_regression_alpha = make_pipeline(PolynomialFeatures(degree=3), Ridge(alpha=1.0))
|
||||
poly_regression_alpha.fit(X_train, y_train)
|
||||
poly_alpha_predictions = poly_regression_alpha.predict(X_test)
|
||||
poly_alpha_mse = mean_squared_error(y_test, poly_alpha_predictions)
|
||||
|
||||
cm = plt.cm.RdBu
|
||||
cm_bright = ListedColormap(['#FF0000', '#0000FF'])
|
||||
# График данных
|
||||
plt.figure(figsize=(10, 6))
|
||||
plt.scatter(X_test[:, 0], X_test[:, 1], c=y_test, cmap='coolwarm')
|
||||
plt.title('Датасет №' + str(ds_cnt))
|
||||
plt.xlabel('X')
|
||||
plt.ylabel('Y')
|
||||
|
||||
h = .02 # шаг регулярной сетки
|
||||
x0_min, x0_max = X[:, 0].min() - .5, X[:, 0].max() + .5
|
||||
x1_min, x1_max = X[:, 1].min() - .5, X[:, 1].max() + .5
|
||||
xx0, xx1 = np.meshgrid(np.arange(x0_min, x0_max, h), np.arange(x1_min, x1_max, h))
|
||||
# График линейной модели
|
||||
plt.figure(figsize=(10, 6))
|
||||
plt.scatter(X_test[:, 0], X_test[:, 1], c=linear_predictions, cmap='coolwarm')
|
||||
plt.title('Линейная ds'+ str(ds_cnt))
|
||||
plt.xlabel('X')
|
||||
plt.ylabel('Y')
|
||||
plt.show()
|
||||
|
||||
Z = clf.decision_function(np.c_[xx0.ravel(), xx1.ravel()]) # 1
|
||||
Z = clf.predict_proba(np.c_[xx0.ravel(), xx1.ravel()])[:, 1] # 2
|
||||
Z = clf.predict(np.c_[xx0.ravel(), xx1.ravel()]) # 3
|
||||
# График полиномиальной модели (degree=3)
|
||||
plt.figure(figsize=(10, 6))
|
||||
plt.scatter(X_test[:, 0], X_test[:, 1], c=poly_predictions, cmap='coolwarm')
|
||||
plt.title('Полиномиальная (degree=3) ds' + str(ds_cnt))
|
||||
plt.xlabel('X')
|
||||
plt.ylabel('Y')
|
||||
plt.show()
|
||||
|
||||
hasattr(clf, "decision_function")
|
||||
# График гребневой модели (degree=3, alpha=1.0)
|
||||
plt.figure(figsize=(10, 6))
|
||||
plt.scatter(X_test[:, 0], X_test[:, 1], c=poly_alpha_predictions, cmap='coolwarm')
|
||||
plt.title('Гребневая (degree=3, alpha=1.0) ds' + str(ds_cnt))
|
||||
plt.xlabel('X')
|
||||
plt.ylabel('Y')
|
||||
plt.show()
|
||||
|
||||
Z = Z.reshape(xx0.shape)
|
||||
current_subplot.contourf(xx0, xx1, Z, cmap=cm, alpha=.8)
|
||||
current_subplot.set_xlim(xx0.min(), xx0.max())
|
||||
current_subplot.set_ylim(xx0.min(), xx1.max())
|
||||
current_subplot.set_xticks(())
|
||||
current_subplot.set_yticks(())
|
||||
current_subplot.set_title(name)
|
||||
current_subplot.text(xx0.max() - .3, xx1.min() + .3, ('%.2f' % score).lstrip('0'),
|
||||
size=15, horizontalalignment='right')
|
||||
# Сравнение качества
|
||||
print('Линейная MSE:', linear_mse)
|
||||
print('Полиномиальная (degree=3) MSE:', poly_mse)
|
||||
print('Гребневая (degree=3, alpha=1.0) MSE:', poly_alpha_mse)
|
||||
|
||||
current_subplot.scatter(X_test[:, 0], X_test[:, 1], c=y_test, cmap=cm_bright, alpha=0.6)
|
||||
|
||||
figure.subplots_adjust(left=.02, right=.98)
|
||||
|
||||
current_subplot.set_title("Input data")
|
||||
current_subplot.text(xx0.max() - .3, xx1.min() + .3, ('%.2f' % score).lstrip('0'), size=15,
|
||||
horizontalalignment='right')
|
||||
|
||||
Loading…
Reference in New Issue
Block a user