机器学习scikit-learn概目
-
线性模型 (linear_model):
1、线性-回归器
① class sklearn.linear_model.LinearRegression()
② class sklearn.linear_model.Ridge()
③ class sklearn.linear_model.Lasso()
④ class sklearn.linear_model.ElasticNet()
2、逻辑-回归器
① class sklearn.linear_model.LogisticRegression()
3、线性-判别器
① class sklearn.discriminat_analysis.LinearDiscriminantAnalysis()
-
决策树 (tree):
1、决策树-回归器
① class sklearn.tree.DecisionTreeRegressor()
2、决策树-分类器
① class sklearn.tree.DecisionTreeClassifier() -
贝叶斯分类 (naive_bays):
1、高斯贝叶斯-分类器
① class sklearn.naive_bayes.GaussianNB()
2、多项式贝叶斯-分类器
① class sklearn.naive_bayes.MultinomialNB()
3、伯努力贝叶斯-分类器
① class sklearn.naive_bayes.BernoulliNB() -
K邻近 (neighbors):
1、K邻近-分类器
① class sklearn.neighbors.KNeighborsClassifier()
2、K邻近-回归器
① class sklearn.neighbors.KNeighborsRegressor() -
降维 (decomposition):
1、主成分分析法
① class sklearn.decomposition.PCA()
② class sklearn.decomposition.IncrementalPCA()
③ class sklearn.decomposition.KernelPCA()
2、流形法 (manifold)
① class sklearn.manifold.MDS()
② class sklearn.manifold.Isomap()
③ class sklearn.manifold.LocallyLinearEmbedding() -
聚类 (cluster):
1、K均值-聚类器
① class sklearn.cluster.KMeans()
2、密度-聚类器
① class sklearn.cluster.DBSCAN()
3、层次-聚类器
① class sklearn.cluster.AgglomerativeClustering() -
支持向量机 (svm):
1、线性-分类器
① class sklearn.svm.LinearSVC()
2、线性-回归器
① class sklearn.svm.LinearSVR()
3、非线性-分类器
① class sklearn.svm.SVC()
4、非线性-回归器
① class sklearn.svm.SVR() -
人工神经网络 (neural_network):
1、多层-分类器
① class sklearn.neural_network.MLPClassifier()
2、多层-回归器
① class sklearn.neural_network.MLPRegressor() -
半监督学习 (semi_supervised):
1、标签传播-回归器
① class sklearn.semi_supervised.LabelPropagation()
② class sklearn.semi_supervised.LabelSpreading() -
集成学习 (ensemble):
1、加法-分类器
① class sklearn.ensemble.AdaBoostClassifier()
2、加法-回归器
① class sklearn.ensemble.AdaBoostRegressor()
3、梯度提升-分类器
① class sklearn.ensemble.GradientBoostingClassifier()
4、梯度提升-回归器
① class sklearn.ensemble.GradientBoostingRegressor()
5、随机森林-分类器
① class sklearn.ensemble.RandomForestClassifier()
6、随机森林-回归器
① class sklearn.ensemble.RandomForestRegressor()
-
数据预处理 (preprocessing):
1、二元化
① class sklearn.preprocessing.Binarizer()
2、独热码
① class sklearn.preprocessing.OneHotEncoder()
3、标准化
① class sklearn.preprocessing.MinMaxScaler()
② class sklearn.preprocessing.MaxAbsScaler()
③ class sklearn.preprocessing.StandardScaler()
4、正则化
① class sklearn.preprocessing.Normalizer() -
特征选择 (feature_selection):
1、方差式-选择器
① class sklearn.feature_selection.VarianceThreshold()
2、指标式-选择器
① class sklearn.feature_selection.SelectKBest()
3、包裹式-选择器
① class sklearn.feature_selection.RFE()
② class sklearn.feature_selection.RFECV()
3、嵌入式-选择器
① class sklearn.feature_selection.SelectFromModel()
② class sklearn.feature_selection.RFECV() -
模型选择 (model_selection):
1、数据切割
① class sklearn.model_selection.train_test_split()
② class sklearn.model_selection.KFold()
③ class sklearn.model_selection.StratifiedKFold()
④ class sklearn.model_selection.LeaveOneOut()
⑤ class sklearn.model_selection.cross_val_score()
2、验证曲线
① class sklearn.model_selection.validation_curve()
3、学习曲线
② class sklearn.model_selection.learning_curve()
4、参数优化
① class sklearn.model_selection.GridSearchCV()
② class sklearn.model_selection.RandomizedSearchCV() -
评价 (metrics):
1、损失函数
① class sklearn.metrics.zero_one_loss()
② class sklearn.metrics.log_loss()
2、准确率 (分类)
① class sklearn.metrics.accuracy_score()
3、查准率 (分类)
① class sklearn.metrics.precision_score()
4、F1值 (分类)
① class sklearn.metrics.f1_score()
5、Fβ值 (分类)
① class sklearn.metrics.fbeta_score()
6、分类报告 (分类)
① class sklearn.metrics.classification_report()
7、混淆矩阵 (分类)
① class sklearn.metrics.confusion_matrix()
8、准-全曲线 (分类)
① class sklearn.metrics.precision_recall_curve()
9、误差绝对值的平均值 (回归)
① class sklearn.metrics.mean_absolute_error()
10、误差平方的平均值 (回归)
① class sklearn.metrics.mean_squared_error()