print("the test error rate is: %f" % (float(errorCount) / m))
if __name__ == "__main__":
import svmMLiA
dataArr, labelArr = svmMLiA.loadDataSet('testSetSVM.txt')
print("dataArr:\n", dataArr)
print("labenlArr:\n", labelArr)
b, alphas = svmMLiA.smoSimple(dataArr, labelArr, 0.6, 0.001, 40)
print("\nb:\n", b, "\nalphas:\n",
alphas[alphas > 0]) #alphas[alphas>0], only can be used in numpy
print("\nshape of alphas>0:\n", shape(alphas[alphas > 0]))
for i in range(100):
if alphas[i] > 0.0:
print(dataArr[i], labelArr[i])
# b,alphas = svmMLiA.smoPK(dataArr,labelArr,0.6,0.001,40)
# print("\nb2:\n", b, "\nalphas2:\n", alphas[alphas > 0]) # alphas[alphas>0], only can be used in numpy
#
# ws=svmMLiA.calcWs(alphas,dataArr,labelArr)
# print("\nws:\n",ws)
# dataMat=mat(dataArr)
# print("datMat[i]: ",dataMat[2]*mat(ws)+b)
# print("LabelArr[i]: ",labelArr[2])
#svmMLiA.testRbf()
svmMLiA.testDigits(('rbf', 20))
Created on 2015. 10. 27.
@author: KimTaeJin
'''
import svmMLiA
from numpy import *
# dataArr, labelArr = svmMLiA.loadDataSet('../data/testSet.txt')
# print(labelArr)
# reload(svmMLiA)
# b, alphas = svmMLiA.smoSimple(dataArr, labelArr, 0.6, 0.001, 40)
#
# print(b)
# print(alphas[alphas>0])
# b, alphas = svmMLiA.smoP(dataArr, labelArr, 0.6, 0.001, 40)
#
# ws = svmMLiA.calcWs(alphas, dataArr, labelArr)
# print(ws)
#
# dataMat = mat(dataArr)
# dataMat[0] * mat(ws) + b
# reload(svmMLiA)
# svmMLiA.testRbf()
# reload(svmMLiA)
svmMLiA.testDigits(('rbf', 20))
import svmMLiA
#from numpy import *
#dataArr, labelArr = svmMLiA.loadDataSet('testSet.txt')
#print labelArr, "\n"
#b, alphas = svmMLiA.smoSimple(dataArr, labelArr, 0.6, 0.001, 40)
#print b
#b, alphas = svmMLiA.smoP(dataArr, labelArr, 0.6, 0.001, 20)
#ws = svmMLiA.calcWs(alphas, dataArr, labelArr)
#print ws
#svmMLiA.testRbf()
#svmMLiA.testRbf(2.1)
svmMLiA.testDigits()
#svmMLiA.testDigits(('rbf',20))
iter: 30 i:17, pairs changed 1
j not moving enough
iteration number: 0
j not moving enough
iteration number: 1
b: matrix([[-3.8486163]])
alphas[alphas>0] = matrix([[0.09313378, 0.27456007, 0.04445935, 0.3232345 ]])
"""
# To see which points of our dataset are support vectors
for i in range(100):
if alphas[i] > 0.0:
print(dataArr[i], labelArr[i])
"""
output:
[4.658191, 3.507396] -1.0
[3.457096, -0.082216] -1.0
[2.893743, -1.643468] -1.0
[6.080573, 0.418886] 1.0
"""
""" optimised SMOP """
b, alphas = SVM.smoP(dataArr, labelArr, 0.6, 0.001, 40)
""" get alpha for supporting vectors """
ws = SVM.calcWs(alphas, dataArr, labelArr)
""" test rbf kernel """
SVM.testRbf()
""" testdigits classification using rbf kernel """
SVM.testDigits(('rbf', 20))
#!/usr/bin/python
# encoding: utf-8
'''
Created on Dec 2, 2015
@author: yanruibo
'''
import svmMLiA
'''
测试四种核函数的SVM
'''
if __name__ == '__main__':
svmMLiA.testDigits(('lin',))
#svmMLiA.testDigits(('rbf',50))
#svmMLiA.testDigits(('polynomial',100))
#svmMLiA.testDigits(('sigmoid',0.00125,0.4))
'''
b, alphas = sl.smoSimple(dataArray, labelArray, 0.6, 0.001, 40)
# print('b = ', b)
print('alphas numbers = ', shape(alphas[alphas > 0]))
for i in range(100):
if alphas[i] > 0.0:
print(dataArray[i], labelArray[i])
b, alphas = sl.smoP(dataArray, labelArray, 0.6, 0.001, 40)
ws = sl.calcWs(alphas, dataArray, labelArray)
print('ws : ', ws)
dataMat = mat(dataArray)
print('first data label : ', dataMat[0] * mat(ws) + b)
print('labelArray[0] = ', labelArray[0])
print('labelArray[1] = ', labelArray[0])
print('labelArray[2] = ', labelArray[0])
'''
# sl.testRbf()
# k1 = 10 to get the smallest test-error
sl.testDigits(('rbf', 10))
import svmMLiA svmMLiA.testDigits()
if alphas[i] > 0.0:
print(dataArr[i], labelArr[i])
# 看下优化之后的
reload(svmMLiA)
b, alphas = svmMLiA.smoP(dataArr, labelArr, 0.6, 0.001, 40)
# 给出W
reload(svmMLiA)
ws = svmMLiA.calcWs(alphas, dataArr, labelArr)
ws
# 查看分类结果
datMat = sp.mat(dataArr)
datMat[0] * sp.mat(ws) + b # 看预测的
labelArr[0] # 看实际的
# 试试看高斯核的
reload(svmMLiA)
svmMLiA.testGauss() # 默认是1.3
svmMLiA.testGauss(k1=0.5)
svmMLiA.testGauss(k1=0.7)
svmMLiA.testGauss(k1=0.9)
svmMLiA.testGauss(k1=1.1)
svmMLiA.testGauss(k1=1.4)
svmMLiA.testGauss(k1=1.6) # 别搞太大
#看看在数字识别上效果如何
reload(svmMLiA)
svmMLiA.testDigits(('Gauss', 10)) # 这个应该是最好的参数了已经
#求出分离超平面
b,alphas=svmMLiA.smoP(dataArr,labelArr,0.6,0.001,40)
ws=svmMLiA.calcWs(alphas,dataArr,labelArr)
#分类决策函数
lf=svmMLiA.classifySVM(ws,b,dataArr[0])
print( lf )
#结束时间
endtime = time.clock()
print( (endtime-starttime) )
#测试svm
svmMLiA.testRbf(0.1)
'''
svmMLiA.testDigits(kTup=('rbf',20))
#!/usr/bin/python
# encoding: utf-8
'''
Created on Dec 2, 2015
@author: yanruibo
'''
import svmMLiA
'''
测试四种核函数的SVM
'''
if __name__ == '__main__':
svmMLiA.testDigits(('lin', ))
#svmMLiA.testDigits(('rbf',50))
#svmMLiA.testDigits(('polynomial',100))
#svmMLiA.testDigits(('sigmoid',0.00125,0.4))
