Python SVM(支持向量机)实现方法完整示例

本文实例讲述了Python SVM(支持向量机)实现方法。分享给大家供大家参考，具体如下：

运行环境

• Pyhton3

• numpy(科学计算包)

• matplotlib(画图所需，不画图可不必)

计算过程

st=>start: 开始
e=>end: 结束
op1=>operation: 读入数据
op2=>operation: 格式化数据
cond=>condition: 是否达到迭代次数
op3=>operation: 寻找超平面分割最小间隔
ccond=>conditon: 数据是否改变
op4=>operation: 输出结果
st->op1->op2->cond
cond(yes)->op4->e
cond(no)->op3

啊，这markdown flow好难用，我决定就画到这吧=。=

输入样例

/* testSet.txt */
3.542485 1.977398 -1
3.018896 2.556416 -1
7.551510 -1.580030 1
2.114999 -0.004466 -1
8.127113 1.274372 1
7.108772 -0.986906 1
8.610639 2.046708 1
2.326297 0.265213 -1
3.634009 1.730537 -1
0.341367 -0.894998 -1
3.125951 0.293251 -1
2.123252 -0.783563 -1
0.887835 -2.797792 -1
7.139979 -2.329896 1
1.696414 -1.212496 -1
8.117032 0.623493 1
8.497162 -0.266649 1
4.658191 3.507396 -1
8.197181 1.545132 1
1.208047 0.213100 -1
1.928486 -0.321870 -1
2.175808 -0.014527 -1
7.886608 0.461755 1
3.223038 -0.552392 -1
3.628502 2.190585 -1
7.407860 -0.121961 1
7.286357 0.251077 1
2.301095 -0.533988 -1
-0.232542 -0.547690 -1
3.457096 -0.082216 -1
3.023938 -0.057392 -1
8.015003 0.885325 1
8.991748 0.923154 1
7.916831 -1.781735 1
7.616862 -0.217958 1
2.450939 0.744967 -1
7.270337 -2.507834 1
1.749721 -0.961902 -1
1.803111 -0.176349 -1
8.804461 3.044301 1
1.231257 -0.568573 -1
2.074915 1.410550 -1
-0.743036 -1.736103 -1
3.536555 3.964960 -1
8.410143 0.025606 1
7.382988 -0.478764 1
6.960661 -0.245353 1
8.234460 0.701868 1
8.168618 -0.903835 1
1.534187 -0.622492 -1
9.229518 2.066088 1
7.886242 0.191813 1
2.893743 -1.643468 -1
1.870457 -1.040420 -1
5.286862 -2.358286 1
6.080573 0.418886 1
2.544314 1.714165 -1
6.016004 -3.753712 1
0.926310 -0.564359 -1
0.870296 -0.109952 -1
2.369345 1.375695 -1
1.363782 -0.254082 -1
7.279460 -0.189572 1
1.896005 0.515080 -1
8.102154 -0.603875 1
2.529893 0.662657 -1
1.963874 -0.365233 -1
8.132048 0.785914 1
8.245938 0.372366 1
6.543888 0.433164 1
-0.236713 -5.766721 -1
8.112593 0.295839 1
9.803425 1.495167 1
1.497407 -0.552916 -1
1.336267 -1.632889 -1
9.205805 -0.586480 1
1.966279 -1.840439 -1
8.398012 1.584918 1
7.239953 -1.764292 1
7.556201 0.241185 1
9.015509 0.345019 1
8.266085 -0.230977 1
8.545620 2.788799 1
9.295969 1.346332 1
2.404234 0.570278 -1
2.037772 0.021919 -1
1.727631 -0.453143 -1
1.979395 -0.050773 -1
8.092288 -1.372433 1
1.667645 0.239204 -1
9.854303 1.365116 1
7.921057 -1.327587 1
8.500757 1.492372 1
1.339746 -0.291183 -1
3.107511 0.758367 -1
2.609525 0.902979 -1
3.263585 1.367898 -1
2.912122 -0.202359 -1
1.731786 0.589096 -1
2.387003 1.573131 -1

代码实现

# -*- coding:utf-8 -*-
#!python3
__author__ = 'Wsine'
from numpy import *
import matplotlib.pyplot as plt
import operator
import time
def loadDataSet(fileName):
 dataMat = []
 labelMat = []
 with open(fileName) as fr:
   for line in fr.readlines():
     lineArr = line.strip().split('\t')
     dataMat.append([float(lineArr[0]), float(lineArr[1])])
     labelMat.append(float(lineArr[2]))
 return dataMat, labelMat
def selectJrand(i, m):
 j = i
 while (j == i):
   j = int(random.uniform(0, m))
 return j
def clipAlpha(aj, H, L):
 if aj > H:
   aj = H
 if L > aj:
   aj = L
 return aj
class optStruct:
 def __init__(self, dataMatIn, classLabels, C, toler):
   self.X = dataMatIn
   self.labelMat = classLabels
   self.C = C
   self.tol = toler
   self.m = shape(dataMatIn)[0]
   self.alphas = mat(zeros((self.m, 1)))
   self.b = 0
   self.eCache = mat(zeros((self.m, 2)))
def calcEk(oS, k):
 fXk = float(multiply(oS.alphas, oS.labelMat).T * (oS.X * oS.X[k, :].T)) + oS.b
 Ek = fXk - float(oS.labelMat[k])
 return Ek
def selectJ(i, oS, Ei):
 maxK = -1
 maxDeltaE = 0
 Ej = 0
 oS.eCache[i] = [1, Ei]
 validEcacheList = nonzero(oS.eCache[:, 0].A)[0]
 if (len(validEcacheList)) > 1:
   for k in validEcacheList:
     if k == i:
       continue
     Ek = calcEk(oS, k)
     deltaE = abs(Ei - Ek)
     if (deltaE > maxDeltaE):
       maxK = k
       maxDeltaE = deltaE
       Ej = Ek
   return maxK, Ej
 else:
   j = selectJrand(i, oS.m)
   Ej = calcEk(oS, j)
 return j, Ej
def updateEk(oS, k):
 Ek = calcEk(oS, k)
 oS.eCache[k] = [1, Ek]
def innerL(i, oS):
 Ei = calcEk(oS, i)
 if ((oS.labelMat[i] * Ei < -oS.tol) and (oS.alphas[i] < oS.C)) or ((oS.labelMat[i] * Ei > oS.tol) and (oS.alphas[i] > 0)):
   j, Ej = selectJ(i, oS, Ei)
   alphaIold = oS.alphas[i].copy()
   alphaJold = oS.alphas[j].copy()
   if (oS.labelMat[i] != oS.labelMat[j]):
     L = max(0, oS.alphas[j] - oS.alphas[i])
     H = min(oS.C, oS.C + oS.alphas[j] - oS.alphas[i])
   else:
     L = max(0, oS.alphas[j] + oS.alphas[i] - oS.C)
     H = min(oS.C, oS.alphas[j] + oS.alphas[i])
   if (L == H):
     # print("L == H")
     return 0
   eta = 2.0 * oS.X[i, :] * oS.X[j, :].T - oS.X[i, :] * oS.X[i, :].T - oS.X[j, :] * oS.X[j, :].T
   if eta >= 0:
     # print("eta >= 0")
     return 0
   oS.alphas[j] -= oS.labelMat[j] * (Ei - Ej) / eta
   oS.alphas[j] = clipAlpha(oS.alphas[j], H, L)
   updateEk(oS, j)
   if (abs(oS.alphas[j] - alphaJold) < 0.00001):
     # print("j not moving enough")
     return 0
   oS.alphas[i] += oS.labelMat[j] * oS.labelMat[i] * (alphaJold - oS.alphas[j])
   updateEk(oS, i)
   b1 = oS.b - Ei - oS.labelMat[i] * (oS.alphas[i] - alphaIold) * oS.X[i, :] * oS.X[i, :].T - oS.labelMat[j] * (oS.alphas[j] - alphaJold) * oS.X[i, :] * oS.X[j, :].T
   b2 = oS.b - Ei - oS.labelMat[i] * (oS.alphas[i] - alphaIold) * oS.X[i, :] * oS.X[j, :].T - oS.labelMat[j] * (oS.alphas[j] - alphaJold) * oS.X[j, :] * oS.X[j, :].T
   if (0 < oS.alphas[i]) and (oS.C > oS.alphas[i]):
     oS.b = b1
   elif (0 < oS.alphas[j]) and (oS.C > oS.alphas[j]):
     oS.b = b2
   else:
     oS.b = (b1 + b2) / 2.0
   return 1
 else:
   return 0
def smoP(dataMatIn, classLabels, C, toler, maxIter, kTup=('lin', 0)):
 """
 输入：数据集, 类别标签, 常数C, 容错率, 最大循环次数
 输出：目标b, 参数alphas
 """
 oS = optStruct(mat(dataMatIn), mat(classLabels).transpose(), C, toler)
 iterr = 0
 entireSet = True
 alphaPairsChanged = 0
 while (iterr < maxIter) and ((alphaPairsChanged > 0) or (entireSet)):
   alphaPairsChanged = 0
   if entireSet:
     for i in range(oS.m):
       alphaPairsChanged += innerL(i, oS)
     # print("fullSet, iter: ％d i:％d, pairs changed ％d" ％ (iterr, i, alphaPairsChanged))
     iterr += 1
   else:
     nonBoundIs = nonzero((oS.alphas.A > 0) * (oS.alphas.A < C))[0]
     for i in nonBoundIs:
       alphaPairsChanged += innerL(i, oS)
       # print("non-bound, iter: ％d i:％d, pairs changed ％d" ％ (iterr, i, alphaPairsChanged))
     iterr += 1
   if entireSet:
     entireSet = False
   elif (alphaPairsChanged == 0):
     entireSet = True
   # print("iteration number: ％d" ％ iterr)
 return oS.b, oS.alphas
def calcWs(alphas, dataArr, classLabels):
 """
 输入：alphas, 数据集, 类别标签
 输出：目标w
 """
 X = mat(dataArr)
 labelMat = mat(classLabels).transpose()
 m, n = shape(X)
 w = zeros((n, 1))
 for i in range(m):
   w += multiply(alphas[i] * labelMat[i], X[i, :].T)
 return w
def plotFeature(dataMat, labelMat, weights, b):
 dataArr = array(dataMat)
 n = shape(dataArr)[0]
 xcord1 = []; ycord1 = []
 xcord2 = []; ycord2 = []
 for i in range(n):
   if int(labelMat[i]) == 1:
     xcord1.append(dataArr[i, 0])
     ycord1.append(dataArr[i, 1])
   else:
     xcord2.append(dataArr[i, 0])
     ycord2.append(dataArr[i, 1])
 fig = plt.figure()
 ax = fig.add_subplot(111)
 ax.scatter(xcord1, ycord1, s=30, c='red', marker='s')
 ax.scatter(xcord2, ycord2, s=30, c='green')
 x = arange(2, 7.0, 0.1)
 y = (-b[0, 0] * x) - 10 / linalg.norm(weights)
 ax.plot(x, y)
 plt.xlabel('X1'); plt.ylabel('X2')
 plt.show()
def main():
 trainDataSet, trainLabel = loadDataSet('testSet.txt')
 b, alphas = smoP(trainDataSet, trainLabel, 0.6, 0.0001, 40)
 ws = calcWs(alphas, trainDataSet, trainLabel)
 print("ws = \n", ws)
 print("b = \n", b)
 plotFeature(trainDataSet, trainLabel, ws, b)
if __name__ == '__main__':
 start = time.clock()
 main()
 end = time.clock()
 print('finish all in ％s' ％ str(end - start))

输出样例

ws =
 [[ 0.65307162]
 [-0.17196128]]
b =
 [[-2.89901748]]
finish all in 2.5683854014099112

绘图方面还存在一些bug。

希望本文所述对大家Python程序设计有所帮助。

来源：https://www.cnblogs.com/wsine/p/5180615.html