#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Wed Oct 18 17:20:19 2017
@author: vicky
"""
'''
SVC参数解释
(1)C: 目标函数的惩罚系数C,用来平衡分类间隔margin和错分样本的,default C = 1.0;
(2)kernel:参数选择有RBF, Linear, Poly, Sigmoid, 默认的是"RBF";
(3)degree:if you choose 'Poly' in param 2, this is effective, degree决定了多项式的最高次幂;
(4)gamma:核函数的系数('Poly', 'RBF' and 'Sigmoid'), 默认是gamma = 1 / n_features;
(5)coef0:核函数中的独立项,'RBF' and 'Poly'有效;
(6)probablity: 可能性估计是否使用(true or false);
(7)shrinking:是否进行启发式;
(8)tol(default = 1e - 3): svm结束标准的精度;
(9)cache_size: 制定训练所需要的内存(以MB为单位);
(10)class_weight: 每个类所占据的权重,不同的类设置不同的惩罚参数C, 缺省的话自适应;
(11)verbose: 跟多线程有关,不大明白啥意思具体;
(12)max_iter: 最大迭代次数,default = 1, if max_iter = -1, no limited;
(13)decision_function_shape : ‘ovo’ 一对一, ‘ovr’ 多对多 or None 无, default=None
(14)random_state :用于概率估计的数据重排时的伪随机数生成器的种子。
ps:7,8,9一般不考虑。
'''
from sklearn.svm import SVC
import numpy as np
X= np.array([[-1,-1],[-2,-1],[1,1],[2,1]])
y = np.array([1,1,2,2])
clf = SVC()
clf.fit(X,y)
clf.predict([[-0.8,-1]])
'''
NuSVC参数
nu:训练误差的一个上界和支持向量的分数的下界。应在间隔(0,1 ]。
其余同SVC
'''
import numpy as np
X = np.array([[-1, -1], [-2, -1], [1, 1], [2, 1]])
y = np.array([1, 1, 2, 2])
from sklearn.svm import NuSVC
clf = NuSVC()
clf.fit(X, y)
clf.predict([[-0.8,-1]])
'''
LinearSVC 参数解释
C:目标函数的惩罚系数C,用来平衡分类间隔margin和错分样本的,default C = 1.0;
loss :指定损失函数
penalty :
dual :选择算法来解决对偶或原始优化问题。当n_samples > n_features 时dual=false。
tol :(default = 1e - 3): svm结束标准的精度;
multi_class:如果y输出类别包含多类,用来确定多类策略, ovr表示一对多,“crammer_singer”优化所有类别的一个共同的目标
如果选择“crammer_singer”,损失、惩罚和优化将会被被忽略。
fit_intercept :
intercept_scaling :
class_weight :对于每一个类别i设置惩罚系数C = class_weight[i]*C,如果不给出,权重自动调整为 n_samples / (n_classes * np.bincount(y))
verbose:跟多线程有关,不大明白啥意思具体<pre name="code" class="python">
'''
from sklearn.svm import SVC
X=[[0],[1],[2],[3]]
Y = [0,1,2,3]
clf = SVC(decision_function_shape='ovo') #ovo为一对一
clf.fit(X,Y)
dec = clf.decision_function([[1]]) #返回的是样本距离超平面的距离
print(dec)
clf.decision_function_shape = "ovr"
dec =clf.decision_function([1]) #返回的是样本距离超平面的距离
print(dec)
#预测
print(clf.predict([1]))
'''
Multi-class classification(多类别分类)
'''
from sklearn.svm import SVC,LinearSVC
X=[[0],[1],[2],[3]]
Y = [0,1,2,3]
'''''
SVC and NuSVC
'''
clf = SVC(decision_function_shape='ovo') #ovo为一对一
clf.fit(X,Y)
print("SVC:",clf.fit(X,Y))
dec = clf.decision_function([[1]]) #返回的是样本距离超平面的距离
print("SVC:",dec)
clf.decision_function_shape = "ovr"
dec =clf.decision_function([1]) #返回的是样本距离超平面的距离
print("SVC:",dec)
#预测
print("预测:",clf.predict([1]))
'''
Unbalanced problems(数据不平衡问题)
'''
import numpy as np
import matplotlib.pyplot as plt
from sklearn import svm
#from sklearn.linear_model import SGDClassifier
# we create 40 separable points
rng = np.random.RandomState(0)
n_samples_1 = 1000
n_samples_2 = 100
X = np.r_[1.5 * rng.randn(n_samples_1, 2),0.5 * rng.randn(n_samples_2, 2) + [2, 2]]
y = [0] * (n_samples_1) + [1] * (n_samples_2)
#print(X)
#print(y)
# fit the model and get the separating hyperplane
clf = svm.SVC(kernel='linear', C=1.0)
clf.fit(X, y)
w = clf.coef_[0]
a = -w[0] / w[1] #a可以理解为斜率
xx = np.linspace(-5, 5)
yy = a * xx - clf.intercept_[0] / w[1] #二维坐标下的直线方程
# get the separating hyperplane using weighted classes
wclf = svm.SVC(kernel='linear', class_weight={1: 10})
wclf.fit(X, y)
ww = wclf.coef_[0]
wa = -ww[0] / ww[1]
wyy = wa * xx - wclf.intercept_[0] / ww[1] #带权重的直线
# plot separating hyperplanes and samples
h0 = plt.plot(xx, yy, 'k-', label='no weights')
h1 = plt.plot(xx, wyy, 'k--', label='with weights')
plt.scatter(X[:, 0], X[:, 1], c=y)
plt.legend()
#不显示线
plt.axis('tight')
plt.show()
'''
regress
'''
from sklearn import svm
X = [[0, 0], [2, 2]]
y = [0.5, 2.5]
clf = svm.SVR()
clf.fit(X, y)
clf.predict([[1, 1]])
'''
线性回归
'''
import numpy as np
from sklearn.svm import SVR
import matplotlib.pyplot as plt
###############################################################################
# Generate sample data
X = np.sort(5 * np.random.rand(40, 1), axis=0) #产生40组数据,每组一个数据,axis=0决定按列排列,=1表示行排列
y = np.sin(X).ravel() #np.sin()输出的是列,和X对应,ravel表示转换成行
###############################################################################
# Add noise to targets
y[::5] += 3 * (0.5 - np.random.rand(8))
###############################################################################
# Fit regression model
svr_rbf = SVR(kernel='rbf', C=1e3, gamma=0.1)
svr_lin = SVR(kernel='linear', C=1e3)
svr_poly = SVR(kernel='poly', C=1e3, degree=2)
y_rbf = svr_rbf.fit(X, y).predict(X)
y_lin = svr_lin.fit(X, y).predict(X)
y_poly = svr_poly.fit(X, y).predict(X)
###############################################################################
# look at the results
lw = 2
plt.scatter(X, y, color='darkorange', label='data')
plt.hold('on')
plt.plot(X, y_rbf, color='navy', lw=lw, label='RBF model')
plt.plot(X, y_lin, color='c', lw=lw, label='Linear model')
plt.plot(X, y_poly, color='cornflowerblue', lw=lw, label='Polynomial model')
plt.xlabel('data')
plt.ylabel('target')
plt.title('Support Vector Regression')
plt.legend()
plt.show()