SVM(support vector machines)

写在开头
　　最近在学习一些关于机器学习的基础算法，结合学习Peter Harrington的《机器学习实战》和李航老师的《统计学习方法》两本书以及网上前辈的笔记，写下了以下的学习过程。
　　代码环境：Pytharm/Python3.7
　　内容有参考也有自己的想法，由于自己的理解不足，文章肯定存在很多错误，还恳请各位批评指正。
　　个人理解的回归就是发现变量之间的关系，也就是求回归系数，经常用回归来预测目标值。回归和分类同属于监督学习，所不同的是回归的目标变量必须是连续数值型。

1.简版SMO算法求分离超平面源码

import random
import numpy
import copy


# -----------------数据处理格式化-------------------
def loadDataSet():
    dataMat = []
    labelMat = []
    lineArray = []
    # read()和readline()是有区别的
    txt = open("testSet.txt", "r", encoding="utf-8").readlines()  # 打开文件
    for line in txt:
        # strip()
        # 方法用于移除字符串头尾指定的字符（默认为空格或换行符）或字符序列。
        # 注意：该方法只能删除开头或是结尾的字符，不能删除中间部分的字符。
        lines = line.strip('\n')
        lineArray = lines.split()  # lineArray返回的是一个含有三个元素的list
        # 进行向量扩充，简化回归，由此计算出的weight[0]就是对应的b.
        dataMat.append([float(lineArray[0]), float(lineArray[1])])  # append()能够添加列表作为元素，区分extend（）
        labelMat.append(int(lineArray[2]))
    return dataMat, labelMat


# 函数情况和书中的叙述不符！
# 书中叙述：i是第一个alpha的下标值，m是所有alpha的数目
# 只要函数值不等于输入值i,函数就会进行随机选择
# 自己还是修改为一下形式
def selectJrand(i, m):
    j = int(random.uniform(0, m))
    if j == i:
        selectJrand(i, m)
    return j


# ------------限定数的范围------------
def clipAlpha(aj, H, L):
    if aj > H:
        aj = H
    if L > aj:
        aj = L
    return aj


# 输入参数依次是：数据集，类别标签，常数C，容错率，推出前最大的循环次数
def smoSimple(dataMatIn, classLabels, C, toler, maxIter):
    dataMatrix = numpy.mat(dataMatIn)
    labelMat = numpy.mat(classLabels).transpose()
    b = 0
    m, n = numpy.shape(dataMatrix)
    alphas = numpy.mat(numpy.zeros((m, 1)))
    iter = 0
    while (iter < maxIter):  # 外循环
        alphaPairsChanged = 0  # 用于记录alpha值是否进行优化
        for i in range(m):  # 内循环
            # 预测类别,multiply是对应元素相乘，不是矩阵相乘。.T代表转置
            # dataMatrix[i, :].T,取第i行元素，并转置
            # fXi是分类决策函数，《统计学习方法》106页公式（线性可分？）
            fXi = float(numpy.multiply(alphas, labelMat).T * (dataMatrix * dataMatrix[i, :].T)) + b
            # 预测误差
            Ei = fXi - float(labelMat[i])  # if checks if an example violates KKT conditions
            # 判断alpha是否需要优化
            if ((labelMat[i] * Ei < -toler) and (alphas[i] < C)) or \
                    ((labelMat[i] * Ei > toler) and (alphas[i] > 0)):
                j = selectJrand(i, m)  # 随机选择第二个alpha（简化SMO算法的第一步）
                fXj = float(numpy.multiply(alphas, labelMat).T * (dataMatrix * dataMatrix[j, :].T)) + b
                Ej = fXj - float(labelMat[j])  # 误差
                # 浅拷贝(copy)：拷贝父对象，不会拷贝对象的内部的子对象，因此子对象任然会随之改变
                alphaIold = alphas[i].copy()
                alphaJold = alphas[j].copy()
                # 保证alpha在0与C之间，《统计学习方法》126页
                # 由于alpha2的最优解必须满足不等式约束条件（在0与C之间），所以取值L<=alpha2<=H
                # L H是alpha2所在的对角线段端点的界
                if (labelMat[i] != labelMat[j]):
                    L = max(0, alphas[j] - alphas[i])
                    H = min(C, C + alphas[j] - alphas[i])
                else:
                    L = max(0, alphas[j] + alphas[i] - C)
                    H = min(C, alphas[j] + alphas[i])
                if L == H:
                    print("L==H")
                    continue
                # 对i进行修改，修改量与j相同，但是方向相反
                # eta是alpha[j]的最优修改量（但是为什么是这样算出来的呢？）
                # 《统计学习方法》127页，计算的eta,但是符号与此处相反，正因为如此，所以下面进行了判断
                # 核函数表示eta = K11+K22-2K12
                eta = 2.0 * dataMatrix[i, :] * dataMatrix[j, :].T \
                      - dataMatrix[i, :] * dataMatrix[i, :].T \
                      - dataMatrix[j, :] * dataMatrix[j, :].T
                if eta >= 0:
                    print("eta>=0")
                    continue  # 不符合条件，跳出本次循环，重新选择alpha

                alphas[j] -= labelMat[j] * (Ei - Ej) / eta  # 《统计学习方法》127页，更新alpha的值
                alphas[j] = clipAlpha(alphas[j], H, L)  # 《统计学习方法》126页，更新alpha的值后同时要限制大小
                # 判断alphas[j]是否只有轻微改变，如果是就认为没有变化跳出for循环
                if (abs(alphas[j] - alphaJold) < 0.00001):
                    print("j not moving enough")
                    continue
                # alphas[i]变化方向与 alphas[j]大小相同，方向相反
                alphas[i] += labelMat[j] * labelMat[i] * (alphaJold - alphas[j])  # update i by the same amount as j
                # the update is in the oppostie direction
                # 在每次完成两个变量的优化之后，都要重新计算阈值B
                # 《统计学习方法》130页
                b1 = b - Ei - labelMat[i] * (alphas[i] - alphaIold) * dataMatrix[i, :] * dataMatrix[i, :].T - labelMat[
                    j] * (alphas[j] - alphaJold) * dataMatrix[i, :] * dataMatrix[j, :].T
                b2 = b - Ej - labelMat[i] * (alphas[i] - alphaIold) * dataMatrix[i, :] * dataMatrix[j, :].T - labelMat[
                    j] * (alphas[j] - alphaJold) * dataMatrix[j, :] * dataMatrix[j, :].T
                if (0 < alphas[i]) and (C > alphas[i]):
                    b = b1
                elif (0 < alphas[j]) and (C > alphas[j]):
                    b = b2
                else:
                    b = (b1 + b2) / 2.0
                alphaPairsChanged += 1
                print("iter: {} i:{}, pairs changed {}".format(iter, i, alphaPairsChanged))
        if (alphaPairsChanged == 0):
            iter += 1
        else:
            iter = 0
        print("iteration number:{}".format(iter))
    return b, alphas


def main():
    dataMat, labelMat = loadDataSet()
    # print(dataMat, labelMat, sep='\n')
    b, alphas = smoSimple(dataMat, labelMat, 0.6, 0.001, 40)
    print(b, alphas[alphas > 0], sep='\n')


if __name__ == '__main__':
    main()

2. 完整版SMO算法

# -------完成版Platt SMO的支持函数----------------
import numpy
import random
import math


# -----------------数据处理格式化-------------------
def loadDataSet():
    dataMat = []
    labelMat = []
    lineArray = []
    # read()和readline()是有区别的
    txt = open("testSet.txt", "r", encoding="utf-8").readlines()  # 打开文件
    for line in txt:
        # strip()
        # 方法用于移除字符串头尾指定的字符（默认为空格或换行符）或字符序列。
        # 注意：该方法只能删除开头或是结尾的字符，不能删除中间部分的字符。
        lines = line.strip('\n')
        lineArray = lines.split()  # lineArray返回的是一个含有三个元素的list
        # 进行向量扩充，简化回归，由此计算出的weight[0]就是对应的b.
        dataMat.append([float(lineArray[0]), float(lineArray[1])])  # append()能够添加列表作为元素，区分extend（）
        labelMat.append(int(lineArray[2]))
    return dataMat, labelMat


# 函数情况和书中的叙述不符！
# 书中叙述：i是第一个alpha的下标值，m是所有alpha的数目
# 只要函数值不等于输入值i,函数就会进行随机选择
# 自己还是修改为一下形式
def selectJrand(i, m):
    j = int(random.uniform(0, m))
    if j == i:
        selectJrand(i, m)
    return j


# ------------限定数的范围------------
def clipAlpha(aj, H, L):
    if aj > H:
        aj = H
    if L > aj:
        aj = L
    return aj


def kernelTrans(X, A, kTup):  # calc the kernel or transform data to a higher dimensional space
    m, n = numpy.shape(X)
    K = numpy.mat(numpy.zeros((m, 1)))
    if kTup[0] == 'lin':
        K = X * A.T  # linear kernel
    elif kTup[0] == 'rbf':
        for j in range(m):
            deltaRow = X[j, :] - A
            K[j] = deltaRow * deltaRow.T
        K = math.exp(K / (-1 * kTup[1] ** 2))  # divide in NumPy is element-wise not matrix like Matlab
    else:
        raise NameError('Houston We Have a Problem -- That Kernel is not recognized')
    return K


class optStruct:
    def __init__(self, dataMatIn, classLabels, C, toler, kTup):  # Initialize the structure with the parameters
        self.X = dataMatIn
        self.labelMat = classLabels
        self.C = C
        self.tol = toler
        self.m = numpy.shape(dataMatIn)[0]
        self.alphas = numpy.mat(numpy.zeros((self.m, 1)))
        self.b = 0
        self.eCache = numpy.mat(numpy.zeros((self.m, 2)))  # first column is valid flag
        self.K = numpy.mat(numpy.zeros((self.m, self.m)))
        for i in range(self.m):
            self.K[:, i] = kernelTrans(self.X, self.X[i, :], kTup)


def calcEk(oS, k):
    fxk = float(numpy.multiply(oS.alphas, oS.labelMat).T * (oS.X * oS.X[k, :].T)) + oS.b
    Ek = fxk - float(oS.labelMat[k])
    return Ek


# 用于选择第二个alpha（内循环）的值
def selectJ(i, oS, Ei):  # this is the second choice -heurstic, and calcs Ej
    maxK = -1
    maxDeltaE = 0
    Ej = 0
    oS.eCache[i] = [1, Ei]  # set valid #choose the alpha that gives the maximum delta E
    validEcacheList = numpy.nonzero(oS.eCache[:, 0].A)[0]
    if (len(validEcacheList)) > 1:
        for k in validEcacheList:  # loop through valid Ecache values and find the one that maximizes delta E
            if k == i:
                continue  # don't calc for i, waste of time
            Ek = calcEk(oS, k)
            deltaE = abs(Ei - Ek)
            if (deltaE > maxDeltaE):
                maxK = k
                maxDeltaE = deltaE
                Ej = Ek
        return maxK, Ej
    else:  # in this case (first time around) we don't have any valid eCache values
        j = selectJrand(i, oS.m)
        Ej = calcEk(oS, j)
    return j, Ej


def updateEk(oS, k):  # after any alpha has changed update the new value in the cache
    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)  # this has been changed from selectJrand
        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.K[i, j] - oS.K[i, i] - oS.K[j, j]  # changed for kernel
        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)  # added this for the Ecache
        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])  # update i by the same amount as j
        updateEk(oS, i)  # added this for the Ecache                    #the update is in the oppostie direction
        b1 = oS.b - Ei - oS.labelMat[i] * (oS.alphas[i] - alphaIold) * oS.K[i, i] - oS.labelMat[j] * (
                oS.alphas[j] - alphaJold) * oS.K[i, j]
        b2 = oS.b - Ej - oS.labelMat[i] * (oS.alphas[i] - alphaIold) * oS.K[i, j] - oS.labelMat[j] * (
                oS.alphas[j] - alphaJold) * oS.K[j, j]
        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)):  # full Platt SMO
    oS = optStruct(numpy.mat(dataMatIn), numpy.mat(classLabels).transpose(), C, toler, kTup)
    iter = 0
    entireSet = True
    alphaPairsChanged = 0
    while (iter < maxIter) and ((alphaPairsChanged > 0) or (entireSet)):
        alphaPairsChanged = 0
        if entireSet:  # go over all
            for i in range(oS.m):
                alphaPairsChanged += innerL(i, oS)
                print("fullSet, iter: {} i:{}, pairs changed {}".format(iter, i, alphaPairsChanged))
            iter += 1
        else:
            nonBoundIs = numpy.nonzero((oS.alphas.A > 0) * (oS.alphas.A < C))[0]
            for i in nonBoundIs:
                alphaPairsChanged += innerL(i, oS)
                print("non-bound, iter: {} i:{}, pairs changed {}".format(iter, i, alphaPairsChanged))
            iter += 1
        if entireSet:
            entireSet = False  # toggle entire set loop
        elif (alphaPairsChanged == 0):
            entireSet = True
        print("iteration number: {}".format(iter))
    return oS.b, oS.alphas


def main():
    dataMat, labelMat = loadDataSet()
    # print(dataMat, labelMat, sep='\n')
    b, alphas = smoP(dataMat, labelMat, 0.6, 0.001, 40)
    print(b, alphas[alphas > 0], sep='\n')


if __name__ == '__main__':
    main()

参考文档：
［1］《机器学习实战》第六章
［2］《统计学习方法》李航