1.基本概述

• Scikit-learn 也简称 Sklearn, 是机器学习领域当中最知名的 python 模块之一.
  Sklearn 包含了很多种机器学习的方式:

• Classification 分类
• Regression 回归
• Clustering 非监督分类
• Dimensionality reduction 数据降维
• Model Selection 模型选择
• Preprocessing 数据预处理

一目了然的流程图

2.入门实践

# utf-8

from sklearn import datasets
from sklearn.cross_validation import train_test_split,cross_val_score
from sklearn.neighbors import KNeighborsClassifier
from sklearn.linear_model import LinearRegression
from sklearn.svm import SVC
from sklearn import preprocessing
from sklearn.learning_curve import validation_curve
import matplotlib.pyplot as plt
import numpy as np

def knn_test():
    '''
    knn实现iris的分类
    '''
    iris = datasets.load_iris()
    iris_x = iris.data
    iris_y = iris.target
    # print(iris_x[:10])
    # print(iris_y[:10])
    x_train, x_test, y_train, y_test = train_test_split(
        iris_x[:100], iris_y[:100], test_size=0.3)
    print(y_train)

    knn = KNeighborsClassifier()
    knn.fit(x_train, y_train)
    print(knn.predict(x_test))
    print(y_test)

def knn_test2():
    '''
    knn实现iris的分类,并且使用交叉验证，并且划分成多次进行交叉验证，得到一个准确度列表scores
    '''
    iris = datasets.load_iris()
    iris_x = iris.data
    iris_y = iris.target
    x_train, x_test, y_train, y_test = train_test_split(iris_x[:100], iris_y[:100], test_size=0.3)
    knn=KNeighborsClassifier(n_neighbors=5)
    scores=cross_val_score(knn,iris_x,iris_y,cv=5,scoring='accuracy')#cross_val_score for classfication
    print(scores)



def linear_test():
    '''
    生成数据，使用linear regression实现回归，画图
    '''
    # boston = datasets.load_boston()
    # x = boston.data
    # y = boston.target
    x, y = datasets.make_regression(n_samples=50, n_features=1, noise=1)
    x_train, x_test, y_train, y_test = train_test_split(
        x[:100], y[:100], test_size=0.3)
    linear = LinearRegression()
    linear.fit(x_train, y_train)
    linear.predict(x[:4])
    print(linear.score(x_test, y_test))
    plt.scatter(x, y)
    plt.show()


def normalization():
    '''
    归一化（正则化）,使用svm进行分类，并画图比较正则化前后的准确率
    '''
    x, y = datasets.make_classification(n_samples=300, n_features=2, n_redundant=0,
                                        n_informative=1, random_state=22, n_clusters_per_class=1, scale=100)
    x_train, x_test, y_train, y_test = train_test_split(
        x[:100], y[:100], test_size=0.3)
    model = SVC()
    model.fit(x_train, y_train)
    score1=model.score(x_test,y_test)
    # print(x[:5], y[:5])
    plt.subplot(121)
    plt.scatter(x[:, 0], x[:, 1], c=y)

    x = preprocessing.scale(x)
    y = preprocessing.scale(y)
    x_train, x_test, y_train, y_test = train_test_split(
        x[:100], y[:100], test_size=0.3)
    model = SVC()
    model.fit(x_train, y_train)
    score2=model.score(x_test,y_test)
    # print(x[:5], y[:5])
    plt.subplot(122)
    plt.scatter(x[:, 0], x[:, 1], c=y)
    print('precision:',score1)
    print('precision:',score2)
    plt.show()


def param_select():
    '''
    选择合适的knn参数k，分别在分类、回归
    '''
    iris = datasets.load_iris()
    x = iris.data
    y = iris.target
    k_range=range(1,30)
    k_scores=[]
    for k in k_range:
        knn=KNeighborsClassifier(n_neighbors=k)
        #loss=-cross_val_score(knn,x,y,cv=10,scoring="mean_squared_error")# for regression
        # k_scores.append(loss.mean())
        scores = cross_val_score(
            knn, x, y, cv=10, scoring='accuracy')  # for classification
        k_scores.append(scores.mean())
    plt.plot(k_range,k_scores)
    plt.xlabel('value of k for knn')
    # plt.ylabel('crowss validated loss')
    plt.ylabel('crowss validated accuracy')
    plt.show()
    

def validation_curve_test():
    '''
    使用validation curve观察学习曲线，此处展示了过拟合的情况
    '''
    digits=datasets.load_digits()
    x=digits.data
    y=digits.target
    model=SVC()
    param_range=np.logspace(-6,-2.3,5)
    train_loss,test_loss=validation_curve(model,x,y,param_name='gamma',param_range=param_range,cv=10,scoring='mean_squared_error')#数据大小，训练曲线、测试曲线
    
    train_loss_mean=-np.mean(train_loss,axis=1)
    test_loss_mean=-np.mean(test_loss,axis=1)
    
    plt.plot(param_range,train_loss_mean,label='train')
    plt.plot(param_range,test_loss_mean,label='cross-validation')
    plt.xlabel('gamma')
    plt.ylabel('Loss')
    plt.legend(loc='best')
    plt.show()




if __name__ == '__main__':
    # knn_test()
    # linear_test()
    # normalization()
    # knn_test2()
    # param_select()
    #  validation_curve_test()

3.部分结果

linear_test
normalization
validation_curve_test

---

作者：Jasonhaven.D
链接：http://www.jianshu.com/u/ed031e432b82
來源：简书
著作权归作者所有。商业转载请联系作者获得授权，非商业转载请注明出处。