手写线性回归

import matplotlib.pyplot as plt
%matplotlib inline
import numpy as np
from sklearn.linear_model import LinearRegression

#一元二次
#f(x) = w1*x**2 + w2*x +b
#二元一次
#f(x1,x2) = w1*x1 + w2*x2 +b
X = np.linspace(0,10,50).reshape(-1,1)
X = np.concatenate([X**2,X],axis = 1)
X.shape

w = np.random.randint(1,10,size=2)
b = np.random.randint(-5,5,size=1)

#矩阵乘法
y = X.dot(w) + b
y.shape

plt.plot(X[:,1],y,color = 'r')

plt.plot(X[:,1],y,color = 'r')
plt.title('w1:%d.w2:%d.b:%d'%(w[0],w[1],b[0]))

#使用sklearn自带的算法，进行预测
lr = LinearRegression()
lr.fit(X,y)
print(lr.coef_,lr.intercept_)

plt.plot(X[:,1],y,color = 'r')
plt.title('w1:%d.w2:%d.b:%d'%(w[0],w[1],b[0]))
#画出点看一下效果
plt.scatter(X[:,1],y,marker = '*')
x = np.linspace(-2,12,100)
plt.plot(x,2*x**2 + 6*x-4,color = 'green')

#手写线性回归
#epoch训练的次数，梯度下降训练多少
def gradient_descent(X,y,lr,epoch,w,b):
    batch = len(X)#一批多少,长度
    
    for i in range(epoch):
        d_loss = 0#损失：自己计算的跟真实的差距
        dw = [0 for _ in range(len(w))]#斜率的梯度
        db = 0#截距的梯度，开始的时候设为0
        for j in range(batch):
            y_ = 0#预测的值，预测方程：f(x) = w1*x1 + w2*x2 + b
            for n in range(len(w)):
                y_ += X[j][n]*w[n]#取出，和对应位置相乘
            y_ += b
            
            d_loss = -(y[j]-y_)#求导
            for n in range(len(w)):
                dw[n] += X[j][n]*d_loss/float(batch)
            db += 1*d_loss/float(batch)
            #更新一下系数和截距，用梯度下降
        for n in range(len(w)):
            w[n] -= dw[n]*lr[n]
        b -= db*lr[0]
    
    return w,b
    

lr = [0.0001,0.001]
w = np.random.randn(2)
b = np.random.randn(1)[0]

w_,b_ = gradient_descent(X,y,lr,500,w,b)
print(w_,b_)

plt.scatter(X[:,1],y,marker = '*')

x = np.linspace(-2,12,100)
f = lambda x:w_[0]*x**2 + w_[1]*x + b_

plt.plot(x,f(x),color = 'g')