python之实战----KNN之手写数字位图

K近邻法分类和回归

#-*- coding=utf-8 -*-import numpy  as npfrom sklearn import datasets,neighborsfrom sklearn.model_selection  import train_test_splitimport matplotlib.pyplot as plt def load_data():    digits=datasets.load_digits()    X_train=digits.data    y_train=digits.target    return train_test_split(X_train,y_train, test_size=0.25,random_state=0,stratify=y_train)def creat_regression_data(n):    '''    在sin(x)基础上添加噪声    '''    X=5*np.random.rand(n,1)    y=np.sin(X).ravel()    y[::5]+=1*(0.5*np.random.rand(int(n/5)))    return train_test_split(X_train,y_train, test_size=0.25,random_state=0,stratify=y_train)def test_KNeighborsClassifier(*data):    X_train,X_test,y_train,y_test=data    cls=neighbors.KNeighborsClassifier()    cls.fit(X_train,y_train)    print('KNeighborsClassifier Train score :%.2f'%cls.score(X_train,y_train))    print('KNeighborsClassifier Test score :%.2f'%cls.score(X_test,y_test))def test_KNeighborsRegressor(*data):    X_train,X_test,y_train,y_test=data    cls=neighbors.KNeighborsRegressor()    cls.fit(X_train,y_train)    print('KNeighborsRegressor Train score :%.2f'%cls.score(X_train,y_train))    print('KNeighborsRegressor Test score :%.2f'%cls.score(X_test,y_test))if __name__=='__main__':    X_train,X_test,y_train,y_test=load_data()    test_KNeighborsClassifier(X_train,X_test,y_train,y_test)    X_train,X_test,y_train,y_test=creat_regression_data(1000)        test_KNeighborsRegressor(X_train,X_test,y_train,y_test)    

结果是:

PS E:\p> python test.pyKNeighborsClassifier Train score :0.99KNeighborsClassifier Test score :0.98KNeighborsRegressor Train score :0.97KNeighborsRegressor Test score :0.94

看看KNN分类K和权重对评分的影响：

#-*- coding=utf-8 -*-import numpy  as npfrom sklearn import datasets,neighborsfrom sklearn.model_selection  import train_test_splitimport matplotlib.pyplot as plt def load_data():    digits=datasets.load_digits()    X_train=digits.data    y_train=digits.target    return train_test_split(X_train,y_train, test_size=0.25,random_state=0,stratify=y_train)def test_KNeighborsClassifier_k_w(*data):    X_train,X_test,y_train,y_test=data    Ks=np.linspace(1,y_train.size,num=100,endpoint=False,dtype='int')#arange数据类型的size    #linspace最初是从MATLAB中学来的，用此来创建等差数列    #endpoint : bool, optional    #If True, stop is the last sample. Otherwise, it is not included. Default is True.    weights=['uniform','distance']    fig=plt.figure()    ax=fig.add_subplot(1,1,1)    for weight in weights:        Train_scores=[]        Test_scores=[]        for K in Ks:            clf=neighbors.KNeighborsClassifier(weights=weight,n_neighbors=K)            clf.fit(X_train,y_train)            Train_scores.append(clf.score(X_train,y_train))            Test_scores.append(clf.score(X_test,y_test))        ax.plot(Ks,Test_scores,label="test score:weight=%s"%weight)        ax.plot(Ks,Train_scores,label="train score:weight=%s"%weight)    ax.legend(loc='best')    ax.set_xlabel("K")    ax.set_ylabel("Score")    ax.set_ylim(0,1.05)    ax.set_title("KNeighborsClassifier_k_w")    plt.show()if __name__=='__main__':    X_train,X_test,y_train,y_test=load_data()    test_KNeighborsClassifier_k_w( X_train,X_test,y_train,y_test)    

#-*- coding=utf-8 -*-import numpy  as npfrom sklearn import datasets,neighborsfrom sklearn.model_selection  import train_test_splitimport matplotlib.pyplot as plt def load_data():    digits=datasets.load_digits()    X_train=digits.data    y_train=digits.target    return train_test_split(X_train,y_train, test_size=0.25,random_state=0,stratify=y_train)def test_KNeighborsClassifier_k_p(*data):    X_train,X_test,y_train,y_test=data    Ks=np.linspace(1,y_train.size,num=100,endpoint=False,dtype='int')#arange数据类型的size    #linspace最初是从MATLAB中学来的，用此来创建等差数列    #endpoint : bool, optional    #If True, stop is the last sample. Otherwise, it is not included. Default is True.    ps=[1,2,10]    fig=plt.figure()    ax=fig.add_subplot(1,1,1)    for p in ps:        Train_scores=[]        Test_scores=[]        for K in Ks:            clf=neighbors.KNeighborsClassifier(p=p,n_neighbors=K)            clf.fit(X_train,y_train)            Train_scores.append(clf.score(X_train,y_train))            Test_scores.append(clf.score(X_test,y_test))        ax.plot(Ks,Test_scores,label="test score:Minkowski-p=%s"%p)#闵可夫斯基        ax.plot(Ks,Train_scores,label="train score:Minkowski-p=%s"%p)    ax.legend(loc='best')    ax.set_xlabel("P")    ax.set_ylabel("Score")    ax.set_ylim(0,1.05)    ax.set_title("KNeighborsClassifier_k_p")    plt.show()if __name__=='__main__':    X_train,X_test,y_train,y_test=load_data()    test_KNeighborsClassifier_k_p( X_train,X_test,y_train,y_test)    

我想说这图画的好慢，电脑性能太菜了。

关于回归图的情况是一样一样的只不过更光滑一些。