管道、Anova和SVM

先后运行带有anova的单变量特征选择，然后用所选特征训练一个SVM，最后使用子管道，可以将拟合系数映射回原始的特征空间。输出：

              precision    recall  f1-score   support

           0       0.75      0.50      0.60         6           1       0.67      1.00      0.80         6           2       0.67      0.80      0.73         5           3       1.00      0.75      0.86         8

    accuracy                           0.76        25   macro avg       0.77      0.76      0.75        25weighted avg       0.79      0.76      0.76        25

[[-0.23912131  0.          0.          0.         -0.3236911   0.   0.          0.          0.          0.          0.          0.   0.10836648  0.          0.          0.          0.          0.   0.          0.        ] [ 0.43878747  0.          0.          0.         -0.51415652  0.   0.          0.          0.          0.          0.          0.   0.04845652  0.          0.          0.          0.          0.   0.          0.        ] [-0.65382998  0.          0.          0.          0.57962856  0.   0.          0.          0.          0.          0.          0.  -0.04736524  0.          0.          0.          0.          0.   0.          0.        ] [ 0.54403412  0.          0.          0.          0.58478491  0.   0.          0.          0.          0.          0.          0.  -0.11344659  0.          0.          0.          0.          0.   0.          0.        ]]

from sklearn import svmfrom sklearn.datasets import make_classificationfrom sklearn.feature_selection import SelectKBest, f_regressionfrom sklearn.pipeline import make_pipelinefrom sklearn.model_selection import train_test_splitfrom sklearn.metrics import classification_report

print(__doc__)

# 导入一些数据进行训练X, y = make_classification(    n_features=20, n_informative=3, n_redundant=0, n_classes=4,    n_clusters_per_class=2)

X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)

# ANOVA SVM-C# 1)方差分析滤镜，采用3种最佳特征anova_filter = SelectKBest(f_regression, k=3)# 2) svmclf = svm.LinearSVC()

anova_svm = make_pipeline(anova_filter, clf)anova_svm.fit(X_train, y_train)y_pred = anova_svm.predict(X_test)print(classification_report(y_test, y_pred))

coef = anova_svm[:-1].inverse_transform(anova_svm['linearsvc'].coef_)print(coef)

脚本的总运行时间：(0分钟0.341秒)估计的内存使用量： 8 MB下载Python源代码:plot_feature_selection_pipeline.py下载Jupyter notebook源代码:plot_feature_selection_pipeline.ipynb由Sphinx-Gallery生成的画廊☆☆☆为方便大家查阅，小编已将scikit-learn学习路线专栏文章统一整理到公众号底部菜单栏，同步更新中，关注公众号，点击左下方“系列文章”，如图：

欢迎大家和我一起沿着scikit-learn文档这条路线，一起巩固机器学习算法基础。(添加微信：mthler，备注：sklearn学习，一起进【sklearn机器学习进步群】开启打怪升级的学习之旅。)