6 支持向量机（Support vector machines）

6.4 Platt SMO算法（Speeding up optimization with the full Platt SMO）

Platt SMO算法优化部分（更新$\mathbf{\alpha}$α）与简化版SMO一致，区别在于优化过程中Platt SMO算法采用启发式算法进行$\alpha$α对-$(\alpha_i, \alpha_j)$(αi​,αj​)的选取。

• 选取$\alpha_i$αi​

无界$\alpha$α：$0 \lt \alpha \lt C$0<α<C，即$\alpha$α还未到达边界处。
在单次传递整个数据集和单次传递非绑定阿尔法之间进行此更改。

• 选取$\alpha_j$αj​

选取能够使步长最大的$\alpha_j$αj​

计算全局误差$E_j$Ej​，然后从中选取能够使步长最大的$\alpha_j$αj​，即$E_i - E_j$Ei​−Ej​

# Listing 6.3 Support functions for full Platt SMO

class optStruct(object):
    
    def __init__(self, dataMatIn, classLabels, C, toler):
        self.X = dataMatIn
        self.labelMat = classLabels
        self.C = C
        self.tol = toler
        self.m = dataMatIn.shape[0]
        self.alphas = np.matrix(np.zeros((self.m, 1)))
        self.b = 0
        # error cache
        self.eCache = np.matrix(np.zeros((self.m, 2)))
        
def calcEk(oS, k):
    fXk = np.multiply(oS.alphas, oS.labelMat).T * \
        (oS.X * oS.X[k, :].T) + oS.b
    Ek = fXk - oS.labelMat[k]
    return Ek
    
# inner-loop heuristic
def selectJ(i, oS, Ei):
    maxK = -1
    maxDeltaE = 0
    Ej = 0
    oS.eCache[i] = [1, Ei]
    
    validEcacheList = np.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)
            # choose j for maximum step size
            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]
    

# Listing 6.4 Full Platt SMO optimization routine

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)):
        
        # second-choice heuristic
        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)
        # updates ecache
        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])
        # updates ecache
        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 - Ej - \
            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
        return 1
    
    else:
        return 0
            

# Listing 6.5 Full Platt SMO outer loop

def smoP(dataMatIn, classLabels, C, toler, maxIter, kTup=("lin", 0)):
    oS = optStruct(np.matrix(dataMatIn),
                   np.matrix(classLabels).transpose(),
                   C, toler)
    iter = 0
    entireSet = True
    alphaPairsChanged = 0
    
    while (iter < maxIter) and ((alphaPairsChanged > 0) or entireSet):
        alphaPairsChanged = 0
        
        if entireSet:
            # go over all values
            for i in range(oS.m):
                alphaPairsChanged += innerL(i, oS)
            print("fullSet, iter: {}, i: {}, pairs changed: {}" \
                  .format(iter, i, alphaPairsChanged))
            iter += 1
        else:
            # go over non-bound values
            nonBoundIs = np.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
        elif alphaPairsChanged == 0:
            entireSet = True
            
        print("iteration number: {} \n ---".format(iter))
        
    return oS.b, oS.alphas
        

while循环退出条件：

（1）迭代次数达到最大；或者（2）遍历训练集，但没有$\alpha$α对被修改。

SVM训练目标：（1）所有样本的余量（margin）至少为1；（2）余量尽可能大。$C$C的作用是在二者之间取得平衡，当$C$C很大时，分类器试图用分隔超平面（separating hyperplane）将所有样本正确分类。

dataArr, labelArr = loadDataSet("testSet.txt")
b, alphas = smoP(dataArr, labelArr, 0.6, 0.001, 40)

L == H
L == H
L == H
j not moving enough
L == H
L == H
L == H
L == H
L == H
L == H
j not moving enough
L == H
L == H
j not moving enough
L == H
L == H
L == H
L == H
L == H
j not moving enough
fullSet, iter: 0, i: 99, pairs changed: 6
iteration number: 1 
 ---
j not moving enough
non-bound, iter: 1, i: 0, pairs changed: 0
j not moving enough
non-bound, iter: 1, i: 3, pairs changed: 0
j not moving enough
non-bound, iter: 1, i: 4, pairs changed: 0
j not moving enough
non-bound, iter: 1, i: 17, pairs changed: 0
j not moving enough
non-bound, iter: 1, i: 18, pairs changed: 0
j not moving enough
non-bound, iter: 1, i: 25, pairs changed: 0
j not moving enough
non-bound, iter: 1, i: 46, pairs changed: 0
non-bound, iter: 1, i: 55, pairs changed: 0
non-bound, iter: 1, i: 94, pairs changed: 0
iteration number: 2 
 ---
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
L == H
L == H
j not moving enough
fullSet, iter: 2, i: 99, pairs changed: 0
iteration number: 3 
 ---

print(b)

# support vectors
support_vector_indices = np.nonzero(alphas)[0]
support_vector_alphas = alphas[support_vector_indices].getA()
print(support_vector_indices)
print(support_vector_alphas)
print(support_vector_alphas.shape)

support_vector_data = np.array([dataArr[idx] for idx in support_vector_indices])
support_vector_labels = np.array([labelArr[idx] for idx in support_vector_indices])

for idx in support_vector_indices:
    print(dataArr[idx], labelArr[idx])
    
weights = np.zeros(support_vector_data.shape[1])
for idx in range(len(support_vector_indices)):
    weights += support_vector_alphas[idx] * support_vector_labels[idx] * \
        support_vector_data[idx, :]
    
print(weights)
print(weights.shape)

[[-2.89901748]]
[ 0  3  4 17 18 25 46 55 94]
[[0.06961952]
 [0.0169055 ]
 [0.0169055 ]
 [0.0272699 ]
 [0.04522972]
 [0.0272699 ]
 [0.0243898 ]
 [0.06140181]
 [0.06140181]]
(9, 1)
[3.542485, 1.977398] -1.0
[2.114999, -0.004466] -1.0
[8.127113, 1.274372] 1.0
[4.658191, 3.507396] -1.0
[8.197181, 1.545132] 1.0
[7.40786, -0.121961] 1.0
[6.960661, -0.245353] 1.0
[6.080573, 0.418886] 1.0
[3.107511, 0.758367] -1.0
[ 0.65307162 -0.17196128]
(2,)

xcord3 = []
ycord3 = []
for idx in support_vector_indices:
    xcord3.append(dataArr[idx][0])
    ycord3.append(dataArr[idx][1])
    
x_max = max(max(xcord1, xcord2))
x_min = min(min(xcord1, xcord2))
y_max = max(max(ycord1, ycord2))
y_min = min(min(ycord1, ycord2))
x = np.linspace(x_min, x_max)
y = (- (weights[0] * x + b) / weights[1]).getA().flatten()

fig = plt.figure()
ax = fig.add_subplot(111)
ax.scatter(xcord1, ycord1, s=30, c="red", marker="s", label="Class 0")
ax.scatter(xcord2, ycord2, s=30, c="green", label="Class 1")
ax.scatter(xcord3, ycord3, s=120, marker="o", c="", edgecolors="blue")
ax.plot(x, y)
ax.set_xlabel("$x_1$")
ax.set_ylabel("$x_2$")
ax.set_xlim(x_min - (x_max + x_min) / 5, x_max + (x_max + x_min) / 5)
ax.set_ylim(y_min - (y_max + y_min), y_max + (y_max + y_min))
ax.legend()
plt.show()

• SVM系数：

$\begin{aligned} f \left( \mathbf{x} \right) = \mathbf{w}^{\mathrm{T}} \mathbf{x} + b = &amp; \sum_{i - 1}^{m} \alpha_i y^{(i)} \left&lt; \mathbf{x}^{(i)},\mathbf{x} \right&gt; + b \\ \downarrow &amp; \\ \mathbf{w} = &amp; \sum_{i - 1}^{m} \alpha_i y^{(i)} \mathbf{x}^{(i)} \end{aligned}$f(x)=wTx+b=↓w=​i−1∑m​αi​y(i)⟨x(i),x⟩+bi−1∑m​αi​y(i)x(i)​

def calcWs(alphas, dataArr, classLabels):
    X = np.matrix(dataArr)
    labelMat = np.matrix(classLabels).transpose()
    m, n = X.shape
    w = np.zeros((n, 1))
    for i in range(m):
        w += np.multiply(alphas[i] * labelMat[i], X[i, :].T)
        
    return w

ws = calcWs(alphas, dataArr, labelArr)
print(ws)

[[ 0.65307162]
 [-0.17196128]]

datMat = np.matrix(dataArr)
print("P0: ")
print(datMat[0] * np.matrix(ws) + b)
print(labelArr[0])
print("---")

print("P1: ")
print(datMat[1] * np.matrix(ws) + b)
print(labelArr[1])
print("---")

print("P2: ")
print(datMat[2] * np.matrix(ws) + b)
print(labelArr[2])
print("---")

P0: 
[[-0.92555695]]
-1.0
---
P1: 
[[-1.36706674]]
-1.0
---
P2: 
[[2.30436336]]
1.0
---

6.5 核：处理复杂数据（Using kernels for more complex data）

6.5.1 利用核将数据映射到高维

当数据线性不可分时，可以尝试利用核（kernel）函数，将其从低维（lower-dimensional）特征空间（feature space）映射到高维特征空间（higher-dimensional）。

由于SVM优化是通过内积（inner products）运算实现的。我们利用核技巧，即用核函数（kernel trick，kernel substation）运算替换内积运算，将数据映射到高维特征空间。

将核函数$K \left( \mathbf{x, y} \right)$K(x,y)替换内积运算，SVM对偶优化问题表示为：

$\max_{\mathbf{\alpha}}W\left(\mathbf{\alpha} \right) = \sum^{m}_{i=1} \alpha_i - \frac{1}{2} \sum^{m}_{i,j=1} \alpha_i \alpha_j y ^{\left(i\right)} y^{\left(j\right)} K \left( \mathbf{x}^{\left(i\right)}, \mathbf{x}^{\left(j\right)} \right)$αmax​W(α)=i=1∑m​αi​−21​i,j=1∑m​αi​αj​y(i)y(j)K(x(i),x(j))
$\text{s.t.}$s.t.
$\begin{aligned} C \geq \alpha_i \geq &amp; 0 \\ \sum^{m}_{i=1} \alpha_i y ^{\left(i\right)} = &amp; 0 \end{aligned}$C≥αi​≥i=1∑m​αi​y(i)=​00​

SVM推理表达式为：

$f \left( \mathbf{x} \right) = \sum_{i - 1}^{m} \alpha_i y^{(i)} K \left( \mathbf{x}^{(i)},\mathbf{x} \right) + b$f(x)=i−1∑m​αi​y(i)K(x(i),x)+b

6.5.2 径向偏置核函数

径向偏置函数（radial basis function，应该是径向基函数（adial basis function））：
- 输入：矢量
- 输出：标量

输出标量与输入向量的距离相关。

高斯核：

$k \left( \mathbf{x, y} \right) = \exp \left( \frac{- \left \| \mathbf{x - y} \right\|^{2}}{2 \sigma^{2} } \right)$k(x,y)=exp(2σ2−∥x−y∥2​)

高斯核将数据映射到无穷维空间（高维空间）中。

• 径向基函数（radial basis function, RBF）是一个实值函数$\phi$ϕ，它的值只取决于输入向量到原点（或其它中心$\mathbf {c}$c）的距离：

$\phi \left( \mathbf{x} \right) = \phi \left( \left\|\mathbf {x} \right\| \right)$ϕ(x)=ϕ(∥x∥)
$\phi \left( \mathbf{x, c} \right) = \phi \left( \left\|\mathbf {x - c} \right\| \right)$ϕ(x,c)=ϕ(∥x−c∥)

任何满足该条件的函数都是径向基函数，范数通常取欧几里德距离。

$r$r：$r = \left\| \mathbf {x} - \mathbf {x}_i \right\|$r=∥x−xi​∥

$\varepsilon$ε：临界半径（critical radius）倒数

• 常用径向基函数类型：

  • 高斯核（Gaussian）：

  $\phi \left( r \right) = \exp \left( - \left( \varepsilon r \right)^2 \right)$ϕ(r)=exp(−(εr)2)

  • 多元二次核（Multiquadric）：

  $\phi \left(r\right) = \sqrt {1 + \left( \varepsilon r \right)^{2}}$ϕ(r)=1+(εr)2​

  • 逆二次核（Inverse quadratic）:

  $\phi \left(r\right)={\dfrac {1}{1+\left(\varepsilon r\right)^{2}}}$ϕ(r)=1+(εr)21​

  • 逆多元二次核（Inverse multiquadric）:

  $\phi \left(r\right)={\dfrac {1}{\sqrt {1+\left(\varepsilon r\right)^{2}}}}$ϕ(r)=1+(εr)2​1​

  • 多重调和样条核（Polyharmonic spline）:

  $\begin{aligned} \phi \left( r \right) &amp;= r^{k}, &amp; k &amp; = 1,3,5, \cdots \\ \phi \left( r \right) &amp;= r^{k} \ln \left( r \right), &amp; k &amp; = 2,4,6, \cdots \end{aligned}$ϕ(r)ϕ(r)​=rk,=rkln(r),​kk​=1,3,5,⋯=2,4,6,⋯​

  • 薄板样条核（Thin plate spline）:

  $\phi \left( r \right) = r^{2} \ln \left( r \right)$ϕ(r)=r2ln(r)

# Listing 6.6 Kernel transformation function

def kernelTrans(X, A, kTup):
    
    m, n = X.shape
    K = np.matrix(np.zeros((m, 1)))
    
    if kTup[0] == "lin":
        K = X * A.T
        
    elif kTup[0] == "rbf":
        for j in range(m):
            deltaRow = X[j, :] - A
            K[j] = deltaRow * deltaRow.T
            
        # element-wise division
        K = np.exp(K / (-1 * kTup[1]**2))
        
    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):
        
        self.X = dataMatIn
        self.labelMat = classLabels
        self.C = C
        self.tol = toler
        self.m = dataMatIn.shape[0]
        self.alphas = np.matrix(np.zeros((self.m, 1)))
        self.b = 0
        self.eCache = np.matrix(np.zeros((self.m, 2)))
        self.K = np.matrix(np.zeros((self.m, self.m)))
        
        for i in range(self.m):
            self.K[:, i] = kernelTrans(self.X, self.X[i, :], kTup)
            

# Listing 6.7 Changes to innerL() and calcEk() needed to user kernels
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)):
        
        # second-choice heuristic
        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.K[i, j] - oS.K[i, i] - oS.K[j, j]
        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)
        # updates ecache
        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])
        # updates ecache
        updateEk(oS, i)
        
        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
        return 1
    
    else:
        return 0

def calcEk(oS, k):
    fXk = np.multiply(oS.alphas, oS.labelMat).T * \
        oS.K[:, k] + oS.b
    Ek = fXk - oS.labelMat[k]
    return Ek



def smoP(dataMatIn, classLabels, C, toler, maxIter, kTup=("lin", 0)):
    oS = optStruct(np.matrix(dataMatIn),
                   np.matrix(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 values
            for i in range(oS.m):
                alphaPairsChanged += innerL(i, oS)
            print("fullSet, iter: {}, i: {}, pairs changed: {}" \
                  .format(iter, i, alphaPairsChanged))
            iter += 1
        else:
            # go over non-bound values
            nonBoundIs = np.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
        elif alphaPairsChanged == 0:
            entireSet = True
            
        print("iteration number: {} \n ---".format(iter))
        
    return oS.b, oS.alphas
        

6.5.3 核测试

参数$\sigma$σ需要人工确定

#　Listing 6.8 Radial bias test function for classifying with a kernel

def testRbf(kl=1.3):
    
    dataArr, labelArr = loadDataSet("testSetRBF.txt")
    
    # training
    b, alphas = smoP(dataArr, labelArr, 200, 0.0001, 10000, ("rbf", kl))
    datMat = np.matrix(dataArr)
    labelMat = np.matrix(labelArr).transpose()
    svInd = np.nonzero(alphas.A > 0)[0]
    # create matrix of support vectors
    sVs = datMat[svInd]
    labelSV = labelMat[svInd]
    
    print("there are {} support vectors".format(sVs.shape[0]))
    
    # training error rate
    m, n = datMat.shape
    errorCount = 0
    for i in range(m):
        kernelEval = kernelTrans(sVs, datMat[i, :], ("rbf", kl))
        predict = kernelEval.T * np.multiply(labelSV, alphas[svInd]) + b
        
        if np.sign(predict) != np.sign(labelArr[i]):
            errorCount += 1
            
    print("the training error rate is: {}".format(errorCount / m))
    
    
    # plot
    xcord1 = []
    ycord1 = []
    xcord2 = []
    ycord2 = []
    for i in range(m):
        if int(labelArr[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", label="Class 0")
    ax.scatter(xcord2, ycord2, s=30, c="green", label="Class 1")
    ax.scatter(sVs[:, 0].A, sVs[:, 1].A, s=120, marker="o", c="", edgecolors="blue", label="svs")
    ax.set_xlabel("$x_1$")
    ax.set_ylabel("$x_2$")
    ax.legend()
    plt.show()

    
    # test error rate
    dataArr, labelArr = loadDataSet("testSetRBF2.txt")
    errorCount = 0
    datMat = np.matrix(dataArr)
    labelMat = np.matrix(labelArr).transpose()
    
    m, n = datMat.shape
    for i in range(m):
        kernelEval = kernelTrans(sVs, datMat[i, :], ("rbf", kl))
        predict = kernelEval.T * np.multiply(labelSV, alphas[svInd]) + b
        
        if np.sign(predict) != np.sign(labelArr[i]):
            errorCount += 1
            
    print("the test error rate is: {}".format(errorCount / m))
    

testRbf()

L == H
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
j not moving enough
L == H
L == H
L == H
fullSet, iter: 0, i: 99, pairs changed: 30
iteration number: 1 
 ---
non-bound, iter: 1, i: 0, pairs changed: 1
j not moving enough
non-bound, iter: 1, i: 1, pairs changed: 1
non-bound, iter: 1, i: 2, pairs changed: 2
j not moving enough
non-bound, iter: 1, i: 3, pairs changed: 2
j not moving enough
non-bound, iter: 1, i: 6, pairs changed: 2
j not moving enough
non-bound, iter: 1, i: 8, pairs changed: 2
j not moving enough
non-bound, iter: 1, i: 10, pairs changed: 2
j not moving enough
non-bound, iter: 1, i: 11, pairs changed: 2
j not moving enough
non-bound, iter: 1, i: 13, pairs changed: 2
j not moving enough
non-bound, iter: 1, i: 14, pairs changed: 2
non-bound, iter: 1, i: 15, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 16, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 18, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 19, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 21, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 27, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 30, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 31, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 33, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 36, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 40, pairs changed: 3
non-bound, iter: 1, i: 41, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 42, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 45, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 48, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 50, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 52, pairs changed: 3
j not moving enough
non-bound, iter: 1, i: 54, pairs changed: 3
non-bound, iter: 1, i: 56, pairs changed: 4
j not moving enough
non-bound, iter: 1, i: 61, pairs changed: 4
non-bound, iter: 1, i: 62, pairs changed: 5
j not moving enough
non-bound, iter: 1, i: 74, pairs changed: 5
iteration number: 2 
 ---
j not moving enough
non-bound, iter: 2, i: 1, pairs changed: 0
j not moving enough
non-bound, iter: 2, i: 2, pairs changed: 0
j not moving enough
non-bound, iter: 2, i: 3, pairs changed: 0
j not moving enough
non-bound, iter: 2, i: 6, pairs changed: 0
non-bound, iter: 2, i: 8, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 10, pairs changed: 1
non-bound, iter: 2, i: 11, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 13, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 14, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 16, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 18, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 19, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 21, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 27, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 30, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 31, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 33, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 36, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 40, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 41, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 42, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 45, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 48, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 50, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 52, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 54, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 56, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 61, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 62, pairs changed: 2
j not moving enough
non-bound, iter: 2, i: 74, pairs changed: 2
iteration number: 3 
 ---
j not moving enough
non-bound, iter: 3, i: 1, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 2, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 3, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 6, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 8, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 10, pairs changed: 0
non-bound, iter: 3, i: 11, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 13, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 14, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 16, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 18, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 19, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 21, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 27, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 30, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 31, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 33, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 36, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 40, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 41, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 42, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 45, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 48, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 50, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 52, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 54, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 56, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 61, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 62, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 74, pairs changed: 0
iteration number: 4 
 ---
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
L == H
L == H
j not moving enough
L == H
j not moving enough
L == H
L == H
L == H
L == H
L == H
L == H
L == H
L == H
L == H
L == H
fullSet, iter: 4, i: 99, pairs changed: 0
iteration number: 5 
 ---
there are 30 support vectors
the training error rate is: 0.13

the test error rate is: 0.15

testRbf(0.1)

L == H
j not moving enough
j not moving enough
j not moving enough
j not moving enough
fullSet, iter: 0, i: 99, pairs changed: 85
iteration number: 1 
 ---
non-bound, iter: 1, i: 0, pairs changed: 1
non-bound, iter: 1, i: 1, pairs changed: 1
non-bound, iter: 1, i: 2, pairs changed: 2
non-bound, iter: 1, i: 3, pairs changed: 3
non-bound, iter: 1, i: 4, pairs changed: 4
non-bound, iter: 1, i: 5, pairs changed: 5
non-bound, iter: 1, i: 6, pairs changed: 6
non-bound, iter: 1, i: 7, pairs changed: 7
non-bound, iter: 1, i: 8, pairs changed: 8
non-bound, iter: 1, i: 9, pairs changed: 9
non-bound, iter: 1, i: 10, pairs changed: 10
non-bound, iter: 1, i: 13, pairs changed: 11
non-bound, iter: 1, i: 14, pairs changed: 12
j not moving enough
non-bound, iter: 1, i: 15, pairs changed: 12
non-bound, iter: 1, i: 16, pairs changed: 13
j not moving enough
non-bound, iter: 1, i: 17, pairs changed: 13
j not moving enough
non-bound, iter: 1, i: 18, pairs changed: 13
non-bound, iter: 1, i: 19, pairs changed: 14
non-bound, iter: 1, i: 20, pairs changed: 15
non-bound, iter: 1, i: 21, pairs changed: 16
non-bound, iter: 1, i: 22, pairs changed: 17
j not moving enough
non-bound, iter: 1, i: 23, pairs changed: 17
non-bound, iter: 1, i: 24, pairs changed: 18
non-bound, iter: 1, i: 25, pairs changed: 19
j not moving enough
non-bound, iter: 1, i: 26, pairs changed: 19
j not moving enough
non-bound, iter: 1, i: 27, pairs changed: 19
j not moving enough
non-bound, iter: 1, i: 28, pairs changed: 19
j not moving enough
non-bound, iter: 1, i: 29, pairs changed: 19
j not moving enough
non-bound, iter: 1, i: 30, pairs changed: 19
j not moving enough
non-bound, iter: 1, i: 31, pairs changed: 19
j not moving enough
non-bound, iter: 1, i: 32, pairs changed: 19
non-bound, iter: 1, i: 33, pairs changed: 20
non-bound, iter: 1, i: 34, pairs changed: 21
j not moving enough
non-bound, iter: 1, i: 36, pairs changed: 21
j not moving enough
non-bound, iter: 1, i: 37, pairs changed: 21
j not moving enough
non-bound, iter: 1, i: 38, pairs changed: 21
non-bound, iter: 1, i: 39, pairs changed: 22
non-bound, iter: 1, i: 40, pairs changed: 23
j not moving enough
non-bound, iter: 1, i: 41, pairs changed: 23
j not moving enough
non-bound, iter: 1, i: 42, pairs changed: 23
j not moving enough
non-bound, iter: 1, i: 43, pairs changed: 23
non-bound, iter: 1, i: 44, pairs changed: 24
non-bound, iter: 1, i: 45, pairs changed: 25
non-bound, iter: 1, i: 46, pairs changed: 26
non-bound, iter: 1, i: 47, pairs changed: 27
non-bound, iter: 1, i: 48, pairs changed: 28
non-bound, iter: 1, i: 50, pairs changed: 29
non-bound, iter: 1, i: 52, pairs changed: 30
non-bound, iter: 1, i: 53, pairs changed: 31
non-bound, iter: 1, i: 54, pairs changed: 32
non-bound, iter: 1, i: 55, pairs changed: 33
non-bound, iter: 1, i: 56, pairs changed: 34
non-bound, iter: 1, i: 57, pairs changed: 35
non-bound, iter: 1, i: 58, pairs changed: 36
non-bound, iter: 1, i: 59, pairs changed: 37
non-bound, iter: 1, i: 61, pairs changed: 38
non-bound, iter: 1, i: 62, pairs changed: 39
non-bound, iter: 1, i: 63, pairs changed: 40
non-bound, iter: 1, i: 64, pairs changed: 41
non-bound, iter: 1, i: 65, pairs changed: 42
non-bound, iter: 1, i: 66, pairs changed: 43
non-bound, iter: 1, i: 67, pairs changed: 44
non-bound, iter: 1, i: 68, pairs changed: 45
non-bound, iter: 1, i: 69, pairs changed: 46
non-bound, iter: 1, i: 70, pairs changed: 47
non-bound, iter: 1, i: 72, pairs changed: 48
non-bound, iter: 1, i: 73, pairs changed: 49
non-bound, iter: 1, i: 74, pairs changed: 50
non-bound, iter: 1, i: 75, pairs changed: 51
non-bound, iter: 1, i: 76, pairs changed: 52
non-bound, iter: 1, i: 77, pairs changed: 53
non-bound, iter: 1, i: 78, pairs changed: 54
non-bound, iter: 1, i: 79, pairs changed: 55
non-bound, iter: 1, i: 81, pairs changed: 56
non-bound, iter: 1, i: 85, pairs changed: 57
non-bound, iter: 1, i: 87, pairs changed: 58
non-bound, iter: 1, i: 88, pairs changed: 59
non-bound, iter: 1, i: 89, pairs changed: 60
non-bound, iter: 1, i: 90, pairs changed: 61
non-bound, iter: 1, i: 91, pairs changed: 62
non-bound, iter: 1, i: 93, pairs changed: 63
j not moving enough
non-bound, iter: 1, i: 97, pairs changed: 63
non-bound, iter: 1, i: 98, pairs changed: 64
non-bound, iter: 1, i: 99, pairs changed: 65
iteration number: 2 
 ---
non-bound, iter: 2, i: 0, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 1, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 2, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 3, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 4, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 5, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 6, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 7, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 8, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 9, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 10, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 12, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 13, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 14, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 15, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 16, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 18, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 19, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 20, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 21, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 22, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 23, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 24, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 25, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 26, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 27, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 28, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 29, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 30, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 31, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 32, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 33, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 34, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 36, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 37, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 38, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 39, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 40, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 41, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 42, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 43, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 44, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 45, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 46, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 47, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 48, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 50, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 52, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 53, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 54, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 55, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 56, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 57, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 58, pairs changed: 1
non-bound, iter: 2, i: 59, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 61, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 62, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 63, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 64, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 65, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 66, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 67, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 69, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 70, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 72, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 73, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 74, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 75, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 76, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 77, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 78, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 79, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 81, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 85, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 87, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 88, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 89, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 90, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 91, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 93, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 97, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 98, pairs changed: 1
j not moving enough
non-bound, iter: 2, i: 99, pairs changed: 1
iteration number: 3 
 ---
non-bound, iter: 3, i: 0, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 1, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 2, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 3, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 4, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 5, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 6, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 7, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 8, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 9, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 10, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 12, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 13, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 14, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 15, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 16, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 18, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 19, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 20, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 21, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 22, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 23, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 24, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 25, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 26, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 27, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 28, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 29, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 30, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 31, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 32, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 33, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 34, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 36, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 37, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 38, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 39, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 40, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 41, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 42, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 43, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 44, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 45, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 46, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 47, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 48, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 50, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 52, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 53, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 54, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 55, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 56, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 57, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 58, pairs changed: 0
non-bound, iter: 3, i: 59, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 61, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 62, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 63, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 64, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 65, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 66, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 67, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 69, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 70, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 72, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 73, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 74, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 75, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 76, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 77, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 78, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 79, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 81, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 85, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 87, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 88, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 89, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 90, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 91, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 93, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 97, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 98, pairs changed: 0
j not moving enough
non-bound, iter: 3, i: 99, pairs changed: 0
iteration number: 4 
 ---
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
L == H
L == H
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
L == H
j not moving enough
fullSet, iter: 4, i: 99, pairs changed: 6
iteration number: 5 
 ---
j not moving enough
non-bound, iter: 5, i: 0, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 1, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 2, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 3, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 4, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 5, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 6, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 7, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 8, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 9, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 10, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 12, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 13, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 14, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 15, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 16, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 18, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 19, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 20, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 21, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 22, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 23, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 24, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 25, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 26, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 27, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 28, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 29, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 30, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 31, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 32, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 33, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 34, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 36, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 37, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 38, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 39, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 40, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 41, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 42, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 43, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 44, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 45, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 46, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 47, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 48, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 50, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 51, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 52, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 53, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 54, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 55, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 56, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 57, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 58, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 59, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 61, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 62, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 63, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 64, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 65, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 66, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 67, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 69, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 70, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 72, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 73, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 74, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 75, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 76, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 77, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 78, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 79, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 81, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 84, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 85, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 87, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 88, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 89, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 90, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 91, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 92, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 93, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 94, pairs changed: 0
non-bound, iter: 5, i: 95, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 97, pairs changed: 0
non-bound, iter: 5, i: 98, pairs changed: 0
j not moving enough
non-bound, iter: 5, i: 99, pairs changed: 0
iteration number: 6 
 ---
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
L == H
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
j not moving enough
fullSet, iter: 6, i: 99, pairs changed: 0
iteration number: 7 
 ---
there are 88 support vectors
the training error rate is: 0.0

the test error rate is: 0.08

6.6 例：回顾手写数字分类（Example: revisiting handwriting classification）

例子：用SVM识别手写数字

1. 收集：提供文本文件。

2. 准备：由二进制图像创建向量

3. 分析：可视化

4. 训练：径向偏置核使用不同的内核和不同的设置，运行SMO算法

5. 测试：计算错误率

6. 使用：

# Listing 6.9 Support vector machine handwriting recognition
import os

def img2vector(filename):
    returnVect = []
    with open(filename) as fr:
        for lineStr in fr.readlines():
            returnVect.extend([int(bit) for bit in lineStr[0 : 32]])
            
    returnVect = np.array(returnVect, ndmin=2)
    return returnVect

def loadImages(dirName):
    hwLabels = []
    trainingFileList  = os.listdir(dirName)
    m = len(trainingFileList)
    trainingMat = np.zeros((m, 1024))
    for i in range(m):
        fileNameStr = trainingFileList[i]
        fileStr = fileNameStr.split('.')[0]
        
        classNumStr = int(fileStr.split('_')[0])
        if classNumStr == 9:
            hwLabels.append(-1)
        else:
            hwLabels.append(1)
            
        trainingMat[i, :] = img2vector("{}/{}".format(dirName, fileNameStr))
        
    return trainingMat, hwLabels

def testDigits(kTup=("rbf", 10)):
    
    # training
    dataArr, labelArr = loadImages("./trainingDigits")
    b, alphas = smoP(dataArr, labelArr, 200, 0.0001, 10000, kTup)
    
    datMat = np.matrix(dataArr)
    labelMat = np.matrix(labelArr).transpose()
    svInd = np.nonzero(alphas.A > 0)[0]
    sVs = datMat[svInd]
    labelSV = labelMat[svInd]
    print("there are {} Support Vectors".format(sVs.shape[0]))
    
    # training error rate
    m, n = datMat.shape
    errorCount = 0
    for i in range(m):
        kernelEval = kernelTrans(sVs, datMat[i, :], kTup)
        predict = kernelEval.T * np.multiply(labelSV, alphas[svInd]) + b
        if np.sign(predict) != np.sign(labelArr[i]):
            errorCount += 1
    print("the training error rate is: {}".format(errorCount / m))
        
    # test error rate
    dataArr, labelArr = loadImages("./testDigits")
    datMat = np.matrix(dataArr)
    labelMat = np.matrix(labelArr).transpose()
    m, n = datMat.shape
    for i in range(m):
        kernelEval = kernelTrans(sVs, datMat[i, :], kTup)
        predict = kernelEval.T * np.multiply(labelSV, alphas[svInd]) + b
        if np.sign(predict) != np.sign(labelArr[i]):
            errorCount += 1
    print("the test error rate is: {}".format(errorCount / m))

testDigits(("rbf", 20))

there are 157 Support Vectors
the training error rate is: 0.0
the test error rate is: 0.012684989429175475

testDigits(("rbf", 10))

there are 298 Support Vectors
the training error rate is: 0.0
the test error rate is: 0.006342494714587738

注意：SVM是二元分类器，且标签值只能为$\pm 1$±1，因此本例中将手写数字识别任务修改为数字9和非9数字分类，并将9的标签设置为-1，其余设置为1。

多分类SVM参考"A Comparison of Methods for Multiclass Support Vector Machines" by C. W. Hus et al。

6.7 总结（Summary）