【Machine Learning】10.逻辑回归的梯度下降

最新推荐文章于 2023-07-31 20:30:58 发布

KiraFenvy

最新推荐文章于 2023-07-31 20:30:58 发布

阅读量579

点赞数

分类专栏： Machine Learning

本文链接：https://blog.csdn.net/m0_51371693/article/details/127025266

版权

Machine Learning 专栏收录该内容

23 篇文章 20 订阅

订阅专栏

逻辑回归的梯度下降

1.导入
2.载入数据
3.逻辑梯度下降 Logistic Gradient Descent
- 梯度下降实现
- - 梯度计算的代码描述
  - 梯度下降的代码描述
4.课后题

在线性回归多项式回归中有的梯度下降，在逻辑回归中同样有。

1.导入

import copy, math
import numpy as np
%matplotlib widget
import matplotlib.pyplot as plt
from lab_utils_common import  dlc, plot_data, plt_tumor_data, sigmoid, compute_cost_logistic
from plt_quad_logistic import plt_quad_logistic, plt_prob
plt.style.use('./deeplearning.mplstyle')

2.载入数据

X_train = np.array([[0.5, 1.5], [1,1], [1.5, 0.5], [3, 0.5], [2, 2], [1, 2.5]])
y_train = np.array([0, 0, 0, 1, 1, 1])

fig,ax = plt.subplots(1,1,figsize=(4,4))
plot_data(X_train, y_train, ax)

ax.axis([0, 4, 0, 3.5])
ax.set_ylabel('$x_1$', fontsize=12)
ax.set_xlabel('$x_0$', fontsize=12)
plt.show()

在这里插入图片描述

3.逻辑梯度下降 Logistic Gradient Descent

回忆一下梯度下降函数
$\begin{align*} &\text{repeat until convergence:} \; \lbrace \\ & \; \; \;w_j = w_j - \alpha \frac{\partial J(\mathbf{w},b)}{\partial w_j} \tag{1} \; & \text{for j := 0..n-1} \\ & \; \; \; \; \;b = b - \alpha \frac{\partial J(\mathbf{w},b)}{\partial b} \\ &\rbrace \end{align*}$

Where each iteration performs simultaneous updates on $w_j$ for all $j$ , where
$\begin{align*} \frac{\partial J(\mathbf{w},b)}{\partial w_j} &= \frac{1}{m} \sum\limits_{i = 0}^{m-1} (f_{\mathbf{w},b}(\mathbf{x}^{(i)}) - y^{(i)})x_{j}^{(i)} \tag{2} \\ \frac{\partial J(\mathbf{w},b)}{\partial b} &= \frac{1}{m} \sum\limits_{i = 0}^{m-1} (f_{\mathbf{w},b}(\mathbf{x}^{(i)}) - y^{(i)}) \tag{3} \end{align*}$

m is the number of training examples in the data set
$f_{\mathbf{w},b}(x^{(i)})$ is the model’s prediction, while $y^{(i)}$ is the target
For a logistic regression model
$\mathbf{w} \cdot \mathbf{x} + b$
$f_{\mathbf{w},b}(x) = g(z)$
where $g (z)$ is the sigmoid function:
$\frac{1}{1+e^{-z}}$

梯度下降实现

梯度下降算法实现有两个组件：

实现上述等式（1）的循环。这是下面的“gradient_desent”
梯度的计算，上述方程（2,3）。这是下面的“compute_gradient_logistic”。

梯度计算的代码描述

对所有 $w_j$ 和 $b$ 执行上面的等式（2）、（3）。

有很多方法可以实现这一点。下文概述如下：

初始化变量以累积dj_dw和dj-db
对于每个示例
- 计算该示例的错误 $g（\mathbf{w}\cdot\mathbf{x}^{（i）}+b）-\mathbf{y}^{i）}$
- 对于本例中的每个输入值 $x{j}^{（i）}$ ，
  - 将错误乘以输入 $x_{j}^{（i）}$ ，并添加到对应的元素dj_dw中。（上面的方程2）
将错误添加到djdb（上面的等式3）
将dj_db和dj-dw除以示例总数（ $m$ ）
注意 $\mathbf{x}^{（i）}$ 在numpyx[i，：]或x[i]中， $x_{j}^{（i）}$ 是x[i，j]

def compute_gradient_logistic(X, y, w, b): 
    """
    Computes the gradient for linear regression 
 
    Args:
      X (ndarray (m,n): Data, m examples with n features
      y (ndarray (m,)): target values
      w (ndarray (n,)): model parameters  
      b (scalar)      : model parameter
    Returns
      dj_dw (ndarray (n,)): The gradient of the cost w.r.t. the parameters w. 
      dj_db (scalar)      : The gradient of the cost w.r.t. the parameter b. 
    """
    m,n = X.shape
    dj_dw = np.zeros((n,))                           #(n,)
    dj_db = 0.

    for i in range(m):
        f_wb_i = sigmoid(np.dot(X[i],w) + b)          #(n,)(n,)=scalar
        err_i  = f_wb_i  - y[i]                       #scalar
        for j in range(n):
            dj_dw[j] = dj_dw[j] + err_i * X[i,j]      #scalar
        dj_db = dj_db + err_i
    dj_dw = dj_dw/m                                   #(n,)
    dj_db = dj_db/m                                   #scalar
        
    return dj_db, dj_dw

代码调用：

X_tmp = np.array([[0.5, 1.5], [1,1], [1.5, 0.5], [3, 0.5], [2, 2], [1, 2.5]])
y_tmp = np.array([0, 0, 0, 1, 1, 1])
w_tmp = np.array([2.,3.])
b_tmp = 1.
dj_db_tmp, dj_dw_tmp = compute_gradient_logistic(X_tmp, y_tmp, w_tmp, b_tmp)
print(f"dj_db: {dj_db_tmp}" )
print(f"dj_dw: {dj_dw_tmp.tolist()}" )

dj_db: 0.49861806546328574
dj_dw: [0.498333393278696, 0.49883942983996693]

梯度下降的代码描述

与之前线性回归时基本相同

def gradient_descent(X, y, w_in, b_in, alpha, num_iters): 
    """
    Performs batch gradient descent
    
    Args:
      X (ndarray (m,n)   : Data, m examples with n features
      y (ndarray (m,))   : target values
      w_in (ndarray (n,)): Initial values of model parameters  
      b_in (scalar)      : Initial values of model parameter
      alpha (float)      : Learning rate
      num_iters (scalar) : number of iterations to run gradient descent
      
    Returns:
      w (ndarray (n,))   : Updated values of parameters
      b (scalar)         : Updated value of parameter 
    """
    # An array to store cost J and w's at each iteration primarily for graphing later
    J_history = []
    w = copy.deepcopy(w_in)  #avoid modifying global w within function
    b = b_in
    
    for i in range(num_iters):
        # Calculate the gradient and update the parameters
        dj_db, dj_dw = compute_gradient_logistic(X, y, w, b)   

        # Update Parameters using w, b, alpha and gradient
        w = w - alpha * dj_dw               
        b = b - alpha * dj_db               
      
        # Save cost J at each iteration
        if i<100000:      # prevent resource exhaustion 
            J_history.append( compute_cost_logistic(X, y, w, b) )

        # Print cost every at intervals 10 times or as many iterations if < 10
        if i% math.ceil(num_iters / 10) == 0:
            print(f"Iteration {i:4d}: Cost {J_history[-1]}   ")
        
    return w, b, J_history         #return final w,b and J history for graphing

代码调用

w_tmp  = np.zeros_like(X_train[0])
b_tmp  = 0.
alph = 0.1
iters = 10000

w_out, b_out, _ = gradient_descent(X_train, y_train, w_tmp, b_tmp, alph, iters) 
print(f"\nupdated parameters: w:{w_out}, b:{b_out}")

Iteration    0: Cost 0.684610468560574   
Iteration 1000: Cost 0.1590977666870457   
Iteration 2000: Cost 0.08460064176930078   
Iteration 3000: Cost 0.05705327279402531   
Iteration 4000: Cost 0.04290759421682   
Iteration 5000: Cost 0.03433847729884557   
Iteration 6000: Cost 0.02860379802212006   
Iteration 7000: Cost 0.02450156960879306   
Iteration 8000: Cost 0.02142370332569295   
Iteration 9000: Cost 0.019030137124109114   

updated parameters: w:[5.28 5.08], b:-14.222409982019837

画出梯度下降的结果

fig,ax = plt.subplots(1,1,figsize=(5,4))
# plot the probability 
plt_prob(ax, w_out, b_out)

# Plot the original data
ax.set_ylabel(r'$x_1$')
ax.set_xlabel(r'$x_0$')   
ax.axis([0, 4, 0, 3.5])
plot_data(X_train,y_train,ax)

# Plot the decision boundary
x0 = -b_out/w_out[1]
x1 = -b_out/w_out[0]
ax.plot([0,x0],[x1,0], c=dlc["dlblue"], lw=1)
plt.show()

在这里插入图片描述

观察过程：
在这里插入图片描述

4.课后题

这里很明显地点出了线性回归和逻辑回归的梯度下降的区别，区别就在于 $f_{\mathbf{w},b}(x)$ 的定义式不同，在逻辑回归这里定义式是

$f_{\mathbf{w},b}(x) = g(z)$ 其中 $\frac{1}{1+e^{-z}}$ ，是使用了sigmoid函数变成0-1过的
而在线性回归，定义式是 $f_{\mathbf{w},b}(\mathbf{x}) = \mathbf{w} \cdot \mathbf{x} + b \tag{2}$