python调用sklearn库实现svr拟合数据

参考的是b站 南方小鱼儿 的代码，参考网址：SVR模型对连续量的预测（SVM）02 - 生成样本数据_哔哩哔哩_bilibili

代码如下

import numpy as np
from sklearn.svm import SVR
import matplotlib.pyplot as plt

X = np.sort(np.random.rand(40,1)*5,axis=0)
y = np.sin(X).ravel()
y[::5] += (0.5 - np.random.rand(8))*0.1

svr_rbf = SVR(kernel='rbf',C=100,gamma=0.1,epsilon=0.01)
svr_rbf.fit(X,y)

'''
C值越大，模型越复杂
'''

svr_lin = SVR(kernel='linear',C=1,gamma=0.1,epsilon=0.01)
svr_lin.fit(X,y)

svr_poly = SVR(kernel='poly',C=1,gamma=0.1,epsilon=0.01,degree=3)
svr_poly.fit(X,y)

svrs = [svr_rbf,svr_lin,svr_poly]
model_labels = ['RBF','Linear','Polynomial']
model_colors = ['m','c','g']

fig,axes = plt.subplots(1,3,figsize=(15,10))
for ix,svr in enumerate(svrs):
    axes[ix].plot(
        X,
        svr.predict(X),
        lw = 3,
        color = model_colors[ix],
        label = f'{model_labels[ix]}model'
        )
    axes[ix].scatter(
        X[svr.support_],
        y[svr.support_],
        s = 30,
        edgecolor = model_colors[ix],
        facecolor = 'none',
        label=f'{model_labels[ix]}support vector'
        )
    #其他的一些点，即排除掉支持向量的点
    axes[ix].scatter(
        X[np.setdiff1d(np.arange(40),svr.support_)],
        y[np.setdiff1d(np.arange(40),svr.support_)],
        s = 30,
        edgecolor = 'k',
        facecolor = 'none',
        label=f'{model_labels[ix]}other data'
        ) 
    
    axes[ix].legend(
        loc = 'upper center',
        bbox_to_anchor = (0.5,1.1)
        )
fig.suptitle('Support Vector Regression',fontsize=15)
plt.show()

最后结果图片如下