scikit-learn的基本用法(五)——交叉验证1

文章作者：Tyan
博客：noahsnail.com  |  CSDN  |  简书

本文主要介绍scikit-learn中的交叉验证。

• Demo 1

import numpy as npfrom sklearn import datasetsfrom sklearn.cross_validation import train_test_splitfrom sklearn.neighbors import KNeighborsClassifierfrom sklearn.cross_validation import cross_val_score # 加载iris数据集iris = datasets.load_iris()# 读取特征X = iris.data# 读取分类标签y = iris.target# 定义分类器knn = KNeighborsClassifier(n_neighbors = 5)# 进行交叉验证数据评估, 数据分为5部分, 每次用一部分作为测试集scores = cross_val_score(knn, X, y, cv = 5, scoring = 'accuracy')# 输出5次交叉验证的准确率print scores

• 结果

[ 0.96666667  1.          0.93333333  0.96666667  1.        ]

• Demo 2

import numpy as npimport matplotlib.pyplot as pltfrom sklearn import datasetsfrom sklearn.cross_validation import train_test_splitfrom sklearn.neighbors import KNeighborsClassifierfrom sklearn.cross_validation import cross_val_score # 确定knn中k的取值# 加载iris数据集iris = datasets.load_iris()# 读取特征X = iris.data# 读取分类标签y = iris.target# 定义knn中k的取值, 0-10k_range = range(1, 30)# 保存k对应的准确率k_scores = []# 计算每个k取值对应的准确率for k in k_range:    # 获得knn分类器    knn = KNeighborsClassifier(n_neighbors = k)    # 对数据进行交叉验证求准确率    scores = cross_val_score(knn, X, y, cv = 10, scoring = 'accuracy')    # 保存交叉验证结果的准确率均值    k_scores.append(scores.mean())# 绘制k取不同值时的准确率变化图像plt.plot(k_range, k_scores)plt.xlabel('K Value in KNN')plt.ylabel('Cross-Validation Mean Accuracy')plt.show()

• 结果