感知机二分类实战

import numpy as np
import matplotlib.pyplot as plt

np.random.seed(0)
num=100

#生成数据集x1,x2,y0/1
#y=0
x1=np.random.normal(6,1,size=(num))
x2=np.random.normal(3,1,size=(num))
y=np.ones(num)
class1=np.array([x1,x2,y])
# class1=class1.reshape(100,3)    注意这里如果是直接reshape的话，后面绘图时，class的数据并未改变
class1.shape

(3, 100)

#y=0
x1=np.random.normal(3,1,size=(num))
x2=np.random.normal(6,1,size=(num))
y=np.ones(num)*(-1)
class0=np.array([x1,x2,y])
# class0=class0.reshape(100,3)
class0.shape

(3, 100)

class1=class1.T
class0=class0.T

plt.scatter(class1[:,0],class1[:,1])
plt.scatter(class0[:,0],class0[:,1],marker='*')

all_data=np.concatenate((class1,class0))
np.random.shuffle(all_data)
all_data.shape

(200, 3)

train_data_x=all_data[:150,:2]
train_data_y=all_data[:150,-1].reshape(150,1)
test_data_x=all_data[150:,:2]
test_data_y=all_data[150:,-1].reshape(50,1)
train_data_x.shape,train_data_y.shape,test_data_y

w=np.zeros((2,1))
T=10
k=0
train_data=np.concatenate((train_data_x,train_data_y),axis=1)
np.dot(w.T,(train_data[0][-1]*train_data[0][:2]).reshape(2,1))

array([[0.]])

train_data_x.shape,train_data_y.shape

((150, 2), (150, 1))

#训练模型
for t in range(T):
    np.random.shuffle(train_data)
    for i in range(len(train_data)):
        #选取第i个样本
        pred=np.dot(w.T,(train_data[i][-1]*train_data[i][:2]).reshape(2,1))[0,0]
        if pred <= 0:
            w=w+(train_data[i][-1]*train_data[i][:2]).reshape(2,1)

w

array([[ 24.56400279],
[-23.68797133]])

plt.scatter(class1[:,0],class1[:,1])
plt.scatter(class0[:,0],class0[:,1],marker='*')
x=np.arange(10)
y=-(w[0]*x)/w[1]
plt.plot(x,y)