ML之UliR：利用非线性回归，梯度下降法(迭代十万次)求出学习参数θ，进而求得Cost函数最优值

ML之UliR：利用非线性回归，梯度下降法(迭代十万次)求出学习参数θ，进而求得Cost函数最优值

目录

输出结果

代码设计

输出结果

更新……

代码设计

  

   

    

     

    
    

     

      import numpy as np  
     
    
   

    

     

    
    

     

      import random 
     
    
   

    

     

    
    

     
 
     
    
   

    

     

    
    

     

      def genData(numPoints,bias,variance):  
     
    
   

    

     

    
    

     

       x = np.zeros(shape=(numPoints,2)) 
     
    
   

    

     

    
    

     

       y = np.zeros(shape=(numPoints))  
     
    
   

    

     

    
    

     
 for i in range(0,numPoints): 
     
    
   

    

     

    
    

     

       x[i][0]=1 
     
    
   

    

     

    
    

     

       x[i][1]=i 
     
    
   

    

     

    
    

     

       y[i]=(i+bias)+random.uniform(0,1)%variance 
     
    
   

    

     

    
    

     
 return x,y
     
    
   

    

     

    
    

     
 
     
    
   

    

     

    
    

     

      def gradientDescent(x,y,theta,alpha,m,numIterations): 
     
    
   

    

     

    
    

     

       xTran = np.transpose(x) 
     
    
   

    

     

    
    

     
 for i in range(numIterations):
     
    
   

    

     

    
    

     

       hypothesis = np.dot(x,theta) 
     
    
   

    

     

    
    

     

       loss = hypothesis-y 
     
    
   

    

     

    
    

     

       cost = np.sum(loss**2)/(2*m) 
     
    
   

    

     

    
    

     

       gradient=np.dot(xTran,loss)/m
     
    
   

    

     

    
    

     

       theta = theta-alpha*gradient 
     
    
   

    

     

    
    

     
 print ("Iteration %d | cost :%f" %(i,cost))
     
    
   

    

     

    
    

     
 return theta
     
    
   

    

     

    
    

     
 
     
    
   

    

     

    
    

     

      x,y = genData(100, 25, 10)  #100行，
     
    
   

    

     

    
    

     

      print ("x:")
     
    
   

    

     

    
    

     

      print (x)
     
    
   

    

     

    
    

     

      print ("y:")
     
    
   

    

     

    
    

     

      print (y)
     
    
   

    

     

    
    

     
 
     
    
   

    

     

    
    

     

      m,n = np.shape(x)
     
    
   

    

     

    
    

     

      n_y = np.shape(y)  
     
    
   

    

     

    
    

     
  
     
    
   

    

     

    
    

     

      print("m:"+str(m)+" n:"+str(n)+" n_y:"+str(n_y))
     
    
   

    

     

    
    

     
  
     
    
   

    

     

    
    

     

      numIterations = 100000 
     
    
   

    

     

    
    

     

      alpha = 0.0005 
     
    
   

    

     

    
    

     

      theta = np.ones(n) 
     
    
   

    

     

    
    

     

      theta= gradientDescent(x, y, theta, alpha, m, numIterations)
     
    
   

    

     

    
    

     

      print(theta)
     
    
  
 

相关文章
ML之UliR：利用非线性回归，梯度下降法(迭代十万次)求出学习参数θ，进而求得Cost函数最优值

文章来源: yunyaniu.blog.csdn.net，作者：一个处女座的程序猿，版权归原作者所有，如需转载，请联系作者。

原文链接：yunyaniu.blog.csdn.net/article/details/78993485