logistic regression中的梯度上升法

1.梯度上升法

def gradascent(datamatin,classlabels):
    datamat=mat(datamatin)
    labelmat=mat(classlabels).transpose()
    m,n=shape(datamat)
    alpha=0.001
    maxcycles=500
    weights=ones((n,1))
    for k in range(maxcycles):
        h=sigmoid(datamat*weights)
        error=(labelmat-h)
        weights=weights+alpha*datamat.transpose()*error
    return weights
    

2.随机梯度上升法（在线学习法）

def stoGradAscent(datamat,classlabels):
    datamat=mat(datamat)
    classlabels=mat(classlabels).transpose()
    m,n=shape(datamat)
    alpha=0.01
    weights=ones((n,1))
    for i in range(m):
        print(datamat[i])
        print(weights)
        h=sigmoid(sum(datamat[i]*weights))
        error=classlabels[i]-h
        
        weights=weights+alpha*error*datamat[i]
    return weights
        

3.改进的随机梯度上升法

def stoGradAscent2(datamat,classlabels,numiter=150):
    datamat=mat(datamat)
    classlabels=mat(classlabels).transpose()
    m,n=shape(datamat)
    weights=ones((n,1))
    dataIndex=range(m)
    for j in range(numiter):
        for i in range(m):
            alpha=4/(1.0+j+i)+0.01
            randIndex=int(random.uniform(0,len(dataIndex)))
            h=sigmoid(sum(datamat[randIndex]*weights))
            error=classlabels[randIndex]-h
            weights=weights+alpha*error*datamat[randIndex]
            del(list(dataIndex)[randIndex])
    return weights