如何用Python构建机器学习模型？-InfoQ



 写点什么

本文，我们将通过 Python 语言包，来构建一些机器学习模型。

构建机器学习模型的模板

该 Notebook 包含了用于创建主要机器学习算法所需的代码模板。在 scikit-learn 中，我们已经准备好了几个算法。只需调整参数，给它们输入数据，进行训练，生成模型，最后进行预测。

1.线性回归

对于线性回归，我们需要从 sklearn 库中导入 linear_model。我们准备好训练和测试数据，然后将预测模型实例化为一个名为线性回归 LinearRegression 算法的对象，它是 linear_model 包的一个类，从而创建预测模型。之后我们利用拟合函数对算法进行训练，并利用得分来评估模型。最后，我们将系数打印出来，用模型进行新的预测。

# Import modulesfrom sklearn import linear_model
# Create training and test subsetsx_train = train_dataset_predictor_variablesy_train = train_dataset_predicted_variable
x_test  = test_dataset_precictor_variables
# Create linear regression objectlinear = linear_model.LinearRegression()
# Train the model with training data and check the scorelinear.fit(x_train, y_train)linear.score(x_train, y_train)
# Collect coefficientsprint('Coefficient: \n', linear.coef_)print('Intercept: \n', linear.intercept_)
# Make predictionspredicted_values = linear.predict(x_test)

复制代码

2.逻辑回归

在本例中，从线性回归到逻辑回归唯一改变的是我们要使用的算法。我们将 LinearRegression 改为 LogisticRegression。

# Import modulesfrom sklearn.linear_model import LogisticRegression
# Create training and test subsetsx_train = train_dataset_predictor_variablesy_train = train_dataset_predicted_variable
x_test  = test_dataset_precictor_variables
# Create logistic regression objectmodel = LogisticRegression()
# Train the model with training data and checking the scoremodel.fit(x_train, y_train)model.score(x_train, y_train)
# Collect coefficientsprint('Coefficient: \n', model.coef_)print('Intercept: \n', model.intercept_)
# Make predictionspredicted_vaues = model.predict(x_teste)

复制代码

3.决策树

我们再次将算法更改为 DecisionTreeRegressor：

# Import modulesfrom sklearn import tree
# Create training and test subsetsx_train = train_dataset_predictor_variablesy_train = train_dataset_predicted_variable
x_test  = test_dataset_precictor_variables
# Create Decision Tree Regressor Objectmodel = tree.DecisionTreeRegressor()
# Create Decision Tree Classifier Objectmodel = tree.DecisionTreeClassifier()
# Train the model with training data and checking the scoremodel.fit(x_train, y_train)model.score(x_train, y_train)
# Make predictionspredicted_values = model.predict(x_test)

复制代码

4.朴素贝叶斯

我们再次将算法更改为 DecisionTreeRegressor：

# Import modulesfrom sklearn.naive_bayes import GaussianNB
# Create training and test subsetsx_train = train_dataset_predictor_variablesy_train = train_dataset_predicted variable
x_test  = test_dataset_precictor_variables
# Create GaussianNB objectmodel = GaussianNB()
# Train the model with training data model.fit(x_train, y_train)
# Make predictionspredicted_values = model.predict(x_test)

复制代码

5.支持向量机

在本例中，我们使用 SVM 库的 SVC 类。如果是 SVR，它就是一个回归函数：

# Import modulesfrom sklearn import svm
# Create training and test subsetsx_train = train_dataset_predictor_variablesy_train = train_dataset_predicted variable
x_test  = test_dataset_precictor_variables
# Create SVM Classifier object  model = svm.svc()
# Train the model with training data and checking the scoremodel.fit(x_train, y_train)model.score(x_train, y_train)
# Make predictionspredicted_values = model.predict(x_test)

复制代码

6.K- 最近邻

在 KneighborsClassifier 算法中，我们有一个超参数叫做 n_neighbors，就是我们对这个算法进行调整。

# Import modulesfrom sklearn.neighbors import KNeighborsClassifier
# Create training and test subsetsx_train = train_dataset_predictor_variablesy_train = train_dataset_predicted variable
x_test  = test_dataset_precictor_variables
# Create KNeighbors Classifier Objects  KNeighborsClassifier(n_neighbors = 6) # default value = 5
# Train the model with training datamodel.fit(x_train, y_train)
# Make predictionspredicted_values = model.predict(x_test)

复制代码

7.K- 均值

# Import modulesfrom sklearn.cluster import KMeans
# Create training and test subsetsx_train = train_dataset_predictor_variablesy_train = train_dataset_predicted variable
x_test  = test_dataset_precictor_variables
# Create KMeans objects k_means = KMeans(n_clusters = 3, random_state = 0)
# Train the model with training datamodel.fit(x_train)
# Make predictionspredicted_values = model.predict(x_test)

复制代码

8.随机森林

# Import modulesfrom sklearn.ensemble import RandomForestClassifier
# Create training and test subsetsx_train = train_dataset_predictor_variablesy_train = train_dataset_predicted variable
x_test  = test_dataset_precictor_variables
# Create Random Forest Classifier objects model = RandomForestClassifier()
# Train the model with training data model.fit(x_train, x_test)
# Make predictionspredicted_values = model.predict(x_test)

复制代码

9.降维

# Import modulesfrom sklearn import decomposition
# Create training and test subsetsx_train = train_dataset_predictor_variablesy_train = train_dataset_predicted variable
x_test  = test_dataset_precictor_variables
# Creating PCA decomposition objectpca = decomposition.PCA(n_components = k)
# Creating Factor analysis decomposition objectfa = decomposition.FactorAnalysis()
# Reduc the size of the training set using PCAreduced_train = pca.fit_transform(train)
# Reduce the size of the training set using PCAreduced_test = pca.transform(test)

复制代码

10.梯度提升和 AdaBoost

# Import modulesfrom sklearn.ensemble import GradientBoostingClassifier
# Create training and test subsetsx_train = train_dataset_predictor_variablesy_train = train_dataset_predicted variable
x_test  = test_dataset_precictor_variables
# Creating Gradient Boosting Classifier objectmodel = GradientBoostingClassifier(n_estimators = 100, learning_rate = 1.0, max_depth = 1, random_state = 0)
# Training the model with training data model.fit(x_train, x_test)
# Make predictionspredicted_values = model.predict(x_test)