def testDigits(kTup=('rbf', 10)):
#导入训练数据
dataArr, labelArr = loadImages('trainingDigits')
b, alphas = svmMLiA2.smoP(dataArr, labelArr, 200, 0.0001, 10000, kTup)
dataMat, labelMat = mat(dataArr), mat(labelArr).transpose()
svInd = nonzero(alphas.A > 0)[0]
sVs = dataMat[svInd]
labelSV = labelMat[svInd]
print("there are %d support vectors" % shape(sVs)[0])
m, n = shape(dataMat)
errorCount = 0
for i in range(m):
kernelEval = svmMLiA2.kernelTrans(sVs, dataMat[i, :], kTup)
predict = kernelEval.T * multiply(labelSV, alphas[svInd]) + b
if sign(predict) != sign(labelArr[i]):
errorCount += 1
print("the training error rate is: %f" % (float(errorCount) / m))
dataArr, labelArr = loadImages('testDigits')
errorCount = 0
dataMat, labelMat = mat(dataArr), mat(labelArr).transpose()
m, n = shape(dataMat)
for i in range(m):
kernelEval = svmMLiA2.kernelTrans(sVs, dataMat[i, :], kTup)
predict = kernelEval.T * multiply(labelSV, alphas[svInd]) + b
if sign(predict) != sign(labelArr[i]):
errorCount += 1
print("the test error rate is: %f" % (float(errorCount) / m))
from numpy import *
import numpy as np
import operator
import svmMLiA2
import matplotlib.pyplot as plt
dataArr, labelArr = svmMLiA2.loadDataSet('nolinear.txt')
# print labelArr
# 主 platt smo 函数
# 数据集:dataArr
# 类别标签:labelArr,
# 错分类系数C: 0.6,
# 容错率:0.001
# 迭代次数: 40
b, alphas = svmMLiA2.smoP(dataArr, labelArr, 0.6, 0.001, 200)
# 根据拉格朗日alphas乘子计算W向量
ws = svmMLiA2.calcWs(alphas, dataArr, labelArr)
print("b:",b)
print("alphas[alphas > 0]:",alphas[alphas > 0])
# 绘制散点图
mydata = mat(dataArr)
# 数据描点
fig = plt.figure()
ax = fig.add_subplot(111)
for i in range(len(mydata)):
if alphas[i]!=0: # KKT条件
ax.scatter(mydata[i,0],mydata[i,1],c='green',marker='s')
elif labelArr[i] == 1:
# -*- coding: utf-8 -*-
# Filename : 03svmRbf.py
from numpy import *
import numpy as np
import operator
import svmMLiA2
import matplotlib.pyplot as plt
k1 = 1.3
# 加载训练集
dataArr, labelArr = svmMLiA2.loadDataSet('testSetRBF.txt')
# 使用Platt SMO分类
# 使用rbf非线性核函数
b, alphas = svmMLiA2.smoP(dataArr, labelArr, 200, 0.0001, 10000,
('rbf', k1)) #C=200 important
datMat = mat(dataArr)
labelMat = mat(labelArr).T
svInd = nonzero(alphas.A > 0)[0]
# 获取支持向量
sVs = datMat[svInd] #get matrix of only support vectors
labelSV = labelMat[svInd]
# 输出支持向量的数量
print "there are %d Support Vectors" % shape(sVs)[0]
print svInd
m, n = shape(datMat)
errorCount = 0
# 计算训练错误率
for i in range(m):
kernelEval = svmMLiA2.kernelTrans(sVs, datMat[i, :], ('rbf', k1))
predict = kernelEval.T * multiply(labelSV, alphas[svInd]) + b