92 lines
3.7 KiB
Python
92 lines
3.7 KiB
Python
|
from random import randrange
|
||
|
|
import numpy as np
|
||
|
|
from sklearn.neural_network import MLPClassifier
|
||
|
|
from sklearn.metrics import accuracy_score, mean_squared_error
|
||
|
|
from matplotlib import pyplot as plt
|
||
|
|
from sklearn.linear_model import LinearRegression
|
||
|
|
from sklearn.model_selection import train_test_split
|
||
|
|
from sklearn.preprocessing import PolynomialFeatures
|
||
|
|
from sklearn.datasets import make_moons
|
||
|
|
from sklearn.pipeline import make_pipeline
|
||
|
|
|
||
|
|
RANDOM_STATE = randrange(50)
|
||
|
|
|
||
|
|
# Генерация случайных данных на основе случайного состояния
|
||
|
|
|
||
|
|
X, y = make_moons(noise=0.3, random_state=RANDOM_STATE)
|
||
|
|
|
||
|
|
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=RANDOM_STATE)
|
||
|
|
|
||
|
|
# Полиномиальная регрессия (3 степень)
|
||
|
|
|
||
|
|
DEGREE = 3
|
||
|
|
|
||
|
|
poly_regression = make_pipeline(PolynomialFeatures(degree=DEGREE), LinearRegression()) # создание модели
|
||
|
|
poly_regression.fit(X_train, y_train) # обучение
|
||
|
|
y_pred_poly_regression = poly_regression.predict(X_test) # предсказание
|
||
|
|
|
||
|
|
# Линейная регрессия
|
||
|
|
|
||
|
|
linear_regression = LinearRegression() # создание модели
|
||
|
|
linear_regression.fit(X_train, y_train) # обучение
|
||
|
|
y_pred_linear_regression = linear_regression.predict(X_test) # предсказание
|
||
|
|
|
||
|
|
# Многослойный персептрон (100 нейронов)
|
||
|
|
|
||
|
|
HIDDEN_LAYER_SIZES = 100
|
||
|
|
ALPHA = 0.01
|
||
|
|
|
||
|
|
perceptron_100 = MLPClassifier(hidden_layer_sizes=(HIDDEN_LAYER_SIZES,), alpha=ALPHA,
|
||
|
|
random_state=RANDOM_STATE) # создание модели
|
||
|
|
perceptron_100.fit(X_train, y_train) # обучение
|
||
|
|
y_pred_perceptron_100 = perceptron_100.predict(X_test) # предсказание
|
||
|
|
|
||
|
|
# Оценка точности и вывод в консоль
|
||
|
|
|
||
|
|
acc_linear_regression = mean_squared_error(y_test, y_pred_linear_regression)
|
||
|
|
acc_poly_regression = mean_squared_error(y_test, y_pred_poly_regression)
|
||
|
|
acc_perceptron_100 = accuracy_score(y_test, y_pred_perceptron_100)
|
||
|
|
|
||
|
|
print(f"Оценка точности: "
|
||
|
|
f"\n Линейная регрессия: {acc_linear_regression}"
|
||
|
|
f"\n Полиномиальная регрессия (3 степень): {acc_poly_regression}"
|
||
|
|
f"\n Многослойный персептрон (100 нейронов): {acc_perceptron_100}")
|
||
|
|
|
||
|
|
# Предсказание классов для точек графика для их визуализации
|
||
|
|
|
||
|
|
x_min, y_min = X[:, 0].min() - 0.5, X[:, 1].min() - 0.5
|
||
|
|
x_max, y_max = X[:, 0].max() + 0.5, X[:, 1].max() + 0.5
|
||
|
|
|
||
|
|
xx, yy = np.meshgrid(np.arange(x_min, x_max, 0.02), np.arange(y_min, y_max, 0.02))
|
||
|
|
|
||
|
|
prediction_data = np.c_[xx.ravel(), yy.ravel()]
|
||
|
|
|
||
|
|
Z_poly_regression = poly_regression.predict(prediction_data)
|
||
|
|
Z_poly_regression = Z_poly_regression.reshape(xx.shape)
|
||
|
|
|
||
|
|
Z_linear_regression = linear_regression.predict(prediction_data)
|
||
|
|
Z_linear_regression = Z_linear_regression.reshape(xx.shape)
|
||
|
|
|
||
|
|
Z_perceptron_100 = perceptron_100.predict(prediction_data)
|
||
|
|
Z_perceptron_100 = Z_perceptron_100.reshape(xx.shape)
|
||
|
|
|
||
|
|
|
||
|
|
# Отрисовка графиков
|
||
|
|
|
||
|
|
def draw_graphic(title, nrows, ncols, index, Z):
|
||
|
|
plt.subplot(nrows, ncols, index)
|
||
|
|
plt.contourf(xx, yy, Z, alpha=0.8)
|
||
|
|
plt.scatter(X_test[:, 0], X_test[:, 1], c=y_test, alpha=0.6)
|
||
|
|
plt.scatter(X_train[:, 0], X_train[:, 1], c=y_train)
|
||
|
|
plt.title(title)
|
||
|
|
plt.xlabel('1 признак')
|
||
|
|
plt.ylabel('2 признак')
|
||
|
|
|
||
|
|
|
||
|
|
draw_graphic('Линейная регрессия', 1, 3, 1, Z_linear_regression)
|
||
|
|
draw_graphic('Полиномиальная регрессия', 1, 3, 2, Z_poly_regression)
|
||
|
|
draw_graphic('Персептрон', 1, 3, 3, Z_perceptron_100)
|
||
|
|
|
||
|
|
plt.tight_layout()
|
||
|
|
plt.show()
|