本文使用sklearn来实现SVM。
首先,导入所需要的包
from sklearn.datasets import load_breast_cancer
from sklearn.svm import SVC
from sklearn.model_selection import train_test_split
import numpy as np
from time import time
import datetime
SVM最重要的就是核函数的选择。一般使用线性核函数、多项式核函数、高斯函数以及sigmoid函数,所以下面给出选择核函数的代码。
def select_kernel(Xtrain, Ytrain):
kernel = ["linear", "poly", "rbf", "sigmoid"]
kernel_dict = {}
for kernel in kernel:
time0 = time() #对现在的时间形成“时间戳”的功能,代表一个时间点
clf = SVC(kernel=kernel
, gamma="auto"
, degree=1#多项式核函数,1就是线性,3就是非线性
, cache_size=5000#允许使用电脑的多少内存来进行计算,单位是MB,默认200MB
).fit(Xtrain, Ytrain)
print("The accuracy under kernel %s is %f" % (kernel, clf.score(Xtest, Ytest)))
kernel_dict_new = {kernel : clf.score(Xtest, Ytest)}
kernel_dict.update(kernel_dict_new)
print(datetime.datetime.fromtimestamp(time()-time0).strftime('%M:%S:%f')) #每个核函数运行花了多长时间
selected_kernel = max(kernel_dict, key=kernel_dict.get)
accuracy = kernel_dict[selected_kernel]
return selected_kernel,accuracy
最后主代码为:
if __name__ == '__main__':
# 导入数据集
data = load_breast_cancer() # data是一个字典
X = data.data # 输出特征矩阵 是一个569行30个特征的数据
y = data.target # 输出标签
no_repeat = np.unique(y) # 有多少个不重复的值
Xtrain, Xtest, Ytrain, Ytest = train_test_split(X, y, test_size=0.3, random_state=420) # 划分训练集和测试集,
selected_kernel,accuracy = select_kernel(Xtrain, Ytrain)
print("select %s accuracy is %f" % (selected_kernel, accuracy))
得到的结果为:
由结果看出,线性核函数使得支持向量机的准确率更高。