逻辑回归及梯度下降（代码）

重点

1 W 计算
2 动画显示（jupyter notebook）
3 数据读取
4 例子数据

---

W迭代计算

def w_cal(xmat,ymat,alpha=0.001,maxIter=101):
    #W init
    W = np.mat(np.random.randn(3,1))
    W_save = []
    for i in range(maxIter):
        H = 1/(1+np.exp(-xmat*W))
        dw = xmat.T*(H-ymat)
        W -= alpha*dw
        W_save.append([W.copy(),i])
    return W,W_save

---

动画显示

%matplotlib inline
from IPython import display

for w in W_save:
    plt.clf()
    w0 = w[0][0,0]
    w1 = w[0][1,0]
    w2 = w[0][2,0]
    plotx1 = np.arange(1,7,0.01)
    plotx2 = -w0/w2-w1/w2*plotx1
    plt.plot(plotx1,plotx2,c='r',label='decision boundary')
    plt.scatter(x[:,1][y==0].A,x[:,2][y==0].A,marker= '^',label="label=0")
    plt.scatter(x[:,1][y==1].A,x[:,2][y==1].A,label="label=0")
    plt.grid()
    plt.legend()
    plt.title('iter:%s'%np.str(w[1]+1))
    #plt.pause(0.1)
    plt.show()
    display.clear_output(wait=True)

---

数据读取

def loadData(filename):
    x = []
    y = []
    with open(filename) as f:
        for line in f.readlines():
            line = line.strip().split()
            x.append([1,float(line[0]),float(line[1])])
            y.append(float(line[-1]))
    xmat = np.mat(x)
    ymat = np.mat(y).T
    return xmat,ymat

---

例子数据
2 1 0
2 2 0
5 4 1
4 5 1
2 3 0
3 2 0
6 5 1
4 1 0
6 3 1
7 4 1