李航《统计学习方法》感知机学习算法实现

import numpy as np

def Train(X_train, Y_train):
    #获取维度参数
    m, n = np.shape(X_train)
    #初始化w，b
    w = np.zeros((n, 1))
    b = 0

    while True:
        count = m     #统计修改次数，若没有变化，则退出while语句
        for i in range(m):
            result = Y_train[i]*(np.dot(X_train[i], w) + b)
            if result <= 0:           #计算yi(w*xi+b)
                count -= 1
                #更新w，b
                for j in range(n):
                    w[j] = w[j] + Y_train[i]*X_train[i][j]
                b = b + Y_train[i]

                print("w:(",w,")","b:",b)
                break

        if count== m:
            break
    return w,b

def main():
    X_train = np.array(([3, 3], [4,3], [1,1]))
    Y_train = np.array(([1,1,-1]))
    print(Train(X_train, Y_train))

if __name__ == '__main__':
    main()

w:( [[ 3.]
 [ 3.]] ) b: 1
w:( [[ 2.]
 [ 2.]] ) b: 0
w:( [[ 1.]
 [ 1.]] ) b: -1
w:( [[ 0.]
 [ 0.]] ) b: -2
w:( [[ 3.]
 [ 3.]] ) b: -1
w:( [[ 2.]
 [ 2.]] ) b: -2
w:( [[ 1.]
 [ 1.]] ) b: -3
(array([[ 1.],
       [ 1.]]), -3)
与统计学习方法中结果一致