代码有什么问题吗

zmllzw

使用sklearn的函数对Diabetes数据集进行加载，并按照自己设定的比例将数据集进行训练集与测试集的划分。线性回归算法对训练集进行训练，并基于测试集进行评估。

1. # 读取相关数据包
   
2. from sklearn import datasets
   
3. from sklearn.model_selection import train_test_split
   
4. import numpy as np
   
5. from sklearn.metrics import r2_score
   
6. import matplotlib.pyplot as plt
   
7. 
   
8. def visualize_feature(X, Y, name_data):
   
9.     plt.figure(figsize=(20, 18))
   
10.     for i in range(len(name_data)):
    
11.         plt.subplot(5, 2, i + 1)
    
12.         plt.scatter(X[:, i], Y[:, i])
    
13.         plt.title(name_data[i])
    
14.     plt.savefig("./Feature_Visualization.jpg")
    
15.     plt.show()
    
16. 
    
17. def create_dataset():
    
18.     # 加载数据集
    
19.     diabetes = datasets.load_diabetes()
    
20.     X = diabetes.data
    
21.     Y = diabetes.target
    
22.     name_data = diabetes.feature_names
    
23.     visualize_feature(X, Y, name_data)
    
24.     # 划分训练集和测试集
    
25.     x_train, x_test, y_train, y_test = train_test_split(X, Y, test_size=0.3, random_state=11)
    
26.     return x_train, y_train, x_test, y_test
    
27. 
    
28. x_train, y_train, x_test, y_test = create_dataset()
    
29. 
    
30. def Regression(x_train, y_train, x_test, y_test):
    
31.     # 初始化参数
    
32.     rng = np.random.RandomState(10)
    
33.     w = rng.randn(1, x_train.shape[1]).reshape(-1, )
    
34.     b = 0
    
35.     w_grad_sum = np.zeros((1, x_train.shape[1])).reshape(-1, )
    
36.     b_grad_sum = np.zeros((1, 1))
    
37.     iteration = 30000
    
38.     lr = 1
    
39.     lv = 0.0001
    
40.     train_acc_list, test_acc_list, epoch_list = [], [], []
    
41. 
    
42.     # 训练和预测
    
43.     for i in range(iteration):
    
44.         w_grad = np.zeros((1, x_train.shape[1])).reshape(-1, )
    
45.         b_grad = np.zeros((1, 1))
    
46.         y = (np.dot(x_train, w) + b).reshape(-1, 1)
    
47. 
    
48.         for j in range(y_train.size):
    
49.             w_grad = w_grad - 2 * (y_train[j] - y[j]) * x_train[j] - 2 * lv * np.sum(w)
    
50.             b_grad = b_grad - 2 * (y_train[j] - y[j])
    
51.         w_grad_sum += w_grad ** 2
    
52.         b_grad_sum += b_grad ** 2
    
53. 
    
54.         w = w - lr * w_grad / (w_grad_sum ** 0.5)
    
55.         b = b - lr * b_grad / (b_grad_sum ** 0.5)
    
56.         if i % 200 == 0:
    
57.             print("-" * 20)
    
58.             train_pred = np.array(np.dot(x_train, w) + b).reshape(-1, 1)
    
59.             test_pred = np.array(np.dot(x_test, w) + b).reshape(-1, 1)
    
60.             score1 = np.clip(r2_score(y_train, train_pred), 0, 100)
    
61.             train_acc_list.append(score1)
    
62.             print("Iteration:[{}/{}] Train acc: {}".format(i, iteration, score1))
    
63.             score2 = np.clip(r2_score(y_test, test_pred), 0, 100)
    
64.             test_acc_list.append(score2)
    
65.             print("Iteration:[{}/{}] Test acc: {}".format(i, iteration, score2))
    
66.             epoch_list.append(i)
    
67. 
    
68.     return train_acc_list, test_acc_list, epoch_list
    
69. 
    
70. 
    
71. # %%
    
72. 
    
73. # 结果可视化
    
74. def visualize_results(epoch, train_acc, test_acc):
    
75.     plt.figure()
    
76.     plt.plot(epoch, train_acc, c='red', label='train')
    
77.     plt.plot(epoch, test_acc, c='blue', label='test')
    
78.     plt.xlabel('Epoch')
    
79.     plt.ylabel('Acc')
    
80.     plt.title('Regression')
    
81.     plt.legend()
    
82.     plt.savefig("./Regression_results.jpg")
    
83.     plt.show()
    
84. 
    
85. 
    
86. if __name__ == '__main__':
    
87.     train_acc, test_acc, epoch = Regression(x_train, y_train, x_test, y_test)
    
88.     visualize_results(epoch, train_acc, test_acc)
    

复制代码