使用 learning_curve 对 load_boston 进行 linear_regression

原创转载请注明出处：https://www.cnblogs.com/agilestyle/p/12697359.html

准备数据

import matplotlib.pyplot as plt
import numpy as np
from sklearn.datasets import load_boston
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import learning_curve
from sklearn.model_selection import train_test_split
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import PolynomialFeatures

# 准备数据
samples = load_boston()

samples

# dict_keys(['data', 'target', 'feature_names', 'DESCR', 'filename'])
samples.keys()
# (506, 13)
samples['data'].shape
# (506,)
samples['target'].shape

# print(samples['DESCR'])

分割训练集和测试集

# 分割训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(samples['data'], samples['target'], test_size=0.2, random_state=3)
# (404, 13)
X_train.shape

建模训练

# 建模训练
lr_model = LinearRegression(normalize=True)
# LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None, normalize=True)
lr_model.fit(X_train, y_train)

评估模型

# 评估模型
train_score = lr_model.score(X_train, y_train)
test_score = lr_model.score(X_test, y_test)
# 0.7239410298290111
train_score
# 0.7952617563243858
test_score

欠拟合，高偏差，构建多项式特征

def polynomial_model(degree=1):
    polynomial_features = PolynomialFeatures(degree=degree, include_bias=False, interaction_only=False)
    linear_regression = LinearRegression(normalize=True)
    pipeline = Pipeline([('polynomial_features', polynomial_features),
                         ('linear_regression', linear_regression)])
    return pipeline


p2_model = polynomial_model(2)
p2_model.fit(X_train, y_train)
train_score = p2_model.score(X_train, y_train)
test_score = p2_model.score(X_test, y_test)

# 0.9305468799409319
print('score on train set: ', train_score)
# 0.8600492818189014
print('score on test set: ', test_score)

绘制学习曲线

学习曲线是用来判断训练模型的一种方法，通过观察绘制出来的学习曲线图，可以比较直观的了解到模型处于一个什么样的状态，如：过拟合（overfitting）或欠拟合（underfitting）。可以很好的表示，当训练数据集增加时，模型对训练数据集拟合的准确性，以及对交叉验证数据集预测的准确性的变化规律。

# array([0.1  , 0.325, 0.55 , 0.775, 1.   ])
train_sizes = np.linspace(0.1, 1.0, 5)
train_sizes, train_scores, test_scores = learning_curve(p2_model, X_train, y_train, cv=10, train_sizes=train_sizes)

train_scores_mean = np.mean(train_scores, axis=1)
test_scores_mean = np.mean(test_scores, axis=1)
plt.grid()

plt.ylim(-1, 1)
plt.plot(train_sizes, train_scores_mean, 'ro-', label="Training score")
plt.plot(train_sizes, test_scores_mean, 'go-', label="Cross-validation score")

plt.legend()

Reference

https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.learning_curve.html

https://blog.csdn.net/qq_36523839/article/details/82556932