机器学习-验证曲线(过拟合与欠拟合的解决)

最新推荐文章于 2025-05-08 08:07:35 发布

原创

最新推荐文章于 2025-05-08 08:07:35 发布 · 1.2k 阅读

3 ·

CC 4.0 BY-SA版权

Santorinisu博客，未经授权，禁止转载!!

文章标签：

#机器学习 #python

验证曲线是解决过拟合和欠拟合的有效工具，通过调整模型参数如逻辑回归中的逆规化参数C，平衡训练和测试集的准确性。在Python机器学习中，找到正则化参数的最佳值能改善模型的泛化能力。实验表明，C参数在[0.01,0.1]区间内能实现较好的平衡，避免过拟合。" 133063348,8750138,华为机试Python：寻找最大树间距问题,"['华为od', 'python', '数据结构', '算法']

Section I: Brief Introduction on ValidatingCurves

Validation curves are a useful tool for improving the performance of a model by addressing issues such as overfitting or underfitting. Validation curves are related to learning curves, but instead of plotting the training and test accuracies as function of the sample size, the values of the model parameters, i.e., the inverse regulation parameter C in logistic regression, are gradually varied, and then the trend of accuracies for both dataset versus regulation strength are plotted.

Personal Views
验证曲线可以用来可视化分析，参数的优化调整对训练/测试数据集的性能影响。

FROM
Sebastian Raschka, Vahid Mirjalili. Python机器学习第二版. 南京：东南大学出版社，2018.

Section II: Code Bundle and Analyses

代码

from sklearn import datasets
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.decomposition import PCA
from sklearn.linear_model import LogisticRegression
from sklearn.pipeline import make_pipeline
import numpy as np
from sklearn.model_selection import validation_curve
import matplotlib.pyplot as plt
import warnings
warnings.filterwarnings("ignore")

plt.rcParams['figure.dpi']=<

最低0.47元/天解锁文章