机器学习--sklearn的常见使用

# 朴素贝叶斯
from sklearn.naive_bayes import GaussianNB
clf = GaussianNB()          #训练模型
clf.fit(features_train,labels_train)     
pred = clf.predit(features_test)        #给出测试变量，预测结果

from sklearn.metrics import accuracy_score
accuracy = accuracy_score(pred, labels_test)   # 预测准确率

#-------------------------------------------------------------------------------------------
# SVM支持向量机，可以通过减少训练集的大小来减少运行时间（牺牲准确率）
from sklearn.svm import SVC

clf = SVC(C=10000,kernel='rbf')     # C的大小决定拟合程度，kernel是核函数，有linear\rbf\polynomial等
t0 = time()                          # C越大越好，但是可能会过拟合
clf.fit(features_train,labels_train)
print "train time is :" , round(time()-t0, 3), "s"    

t0 = time()
pred = clf.predict(features_test)
print "fit time is :" , round(time()-t0,3), "s"
#print pred[10],pred[26],pred[50]               # 得出第10/26/50的预测结果

from sklearn.metrics import accuracy_score
acc = accuracy_score(pred, labels_test)
print acc                                      # 准确率

#--------------------------------------------------------------------------------------------