机器学习实验之肿瘤预测与分析(神经网络)

肿瘤预测与分析(神经网络)

【实验内容】

基于威斯康星乳腺癌数据集,搭建BP神经网络,实现肿瘤预测与分析。

【实验要求】

1.加载sklearn自带的数据集,探索数据。

2.划分训练集与测试集。

3.建立BP模型(评估后可进行调参,从而选择最优参数)。

4.进行模型训练。

5.进行模型预测,对真实数据和预测数据进行可视化(用Axes3D绘制3d散点图)。

6.进行模型评估,并进行预测结果指标统计(统计每一类别的预测准确率、召回率、F1分数)。

7.计算混淆矩阵,并用热力图显示。

注:混淆矩阵(confusion matrix)衡量的是一个分类器分类的准确程度。

   混淆矩阵的每一列代表了预测类别 ,每一列的总数表示预测为该类别的数据的数目;每一行代表了数据的真实归属类别 ,每一行的数据总数表示该类别的数据实例的数目。
from sklearn.datasets import fetch_openml  #导入数据集获取工具
from sklearn.preprocessing import StandardScaler
from sklearn.datasets import load_breast_cancer
from sklearn.utils import check_random_state
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split

加载sklearn自带的数据集,探索数据。

cancers = load_breast_cancer()
cancers
{'data': array([[1.799e+01, 1.038e+01, 1.228e+02, ..., 2.654e-01, 4.601e-01,
         1.189e-01],
        [2.057e+01, 1.777e+01, 1.329e+02, ..., 1.860e-01, 2.750e-01,
         8.902e-02],
        [1.969e+01, 2.125e+01, 1.300e+02, ..., 2.430e-01, 3.613e-01,
         8.758e-02],
        ...,
        [1.660e+01, 2.808e+01, 1.083e+02, ..., 1.418e-01, 2.218e-01,
         7.820e-02],
        [2.060e+01, 2.933e+01, 1.401e+02, ..., 2.650e-01, 4.087e-01,
         1.240e-01],
        [7.760e+00, 2.454e+01, 4.792e+01, ..., 0.000e+00, 2.871e-01,
         7.039e-02]]),
 'target': array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1,
        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0,
        0, 0, 1, 0, 1, 1, 1, 1, 1, 0, 0, 1, 0, 0, 1, 1, 1, 1, 0, 1, 0, 0,
        1, 1, 1, 1, 0, 1, 0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 0, 1, 0, 0, 0,
        1, 1, 1, 0, 1, 1, 0, 0, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 0, 1,
        1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 0, 0, 1, 1, 1, 0, 0, 1, 0, 1, 0,
        0, 1, 0, 0, 1, 1, 0, 1, 1, 0, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1,
        1, 1, 0, 1, 1, 1, 1, 0, 0, 1, 0, 1, 1, 0, 0, 1, 1, 0, 0, 1, 1, 1,
        1, 0, 1, 1, 0, 0, 0, 1, 0, 1, 0, 1, 1, 1, 0, 1, 1, 0, 0, 1, 0, 0,
        0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 1, 1, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0,
        1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 0, 0, 1, 0, 1, 1,
        1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
        0, 0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 1, 0, 0, 1, 1,
        1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 0, 1, 0, 1, 1, 1, 1, 1,
        1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1, 1, 1, 1, 0, 0,
        0, 1, 1, 1, 1, 0, 1, 0, 1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 0,
        0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 0, 0, 0, 1, 0, 0,
        1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, 1, 0, 0, 1, 1,
        1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 1, 0,
        1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0, 1, 0, 1, 1, 1, 1,
        1, 0, 1, 1, 0, 1, 0, 1, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0,
        1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1,
        1, 1, 1, 0, 1, 0, 1, 1, 0, 1, 1, 1, 1, 1, 0, 0, 1, 0, 1, 0, 1, 1,
        1, 1, 1, 0, 1, 1, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1,
        1, 1, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
        1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1]),
 'frame': None,
 'target_names': array(['malignant', 'benign'], dtype='<U9'),
 'DESCR': '.. _breast_cancer_dataset:\n\nBreast cancer wisconsin (diagnostic) dataset\n--------------------------------------------\n\n**Data Set Characteristics:**\n\n    :Number of Instances: 569\n\n    :Number of Attributes: 30 numeric, predictive attributes and the class\n\n    :Attribute Information:\n        - radius (mean of distances from center to points on the perimeter)\n        - texture (standard deviation of gray-scale values)\n        - perimeter\n        - area\n        - smoothness (local variation in radius lengths)\n        - compactness (perimeter^2 / area - 1.0)\n        - concavity (severity of concave portions of the contour)\n        - concave points (number of concave portions of the contour)\n        - symmetry\n        - fractal dimension ("coastline approximation" - 1)\n\n        The mean, standard error, and "worst" or largest (mean of the three\n        worst/largest values) of these features were computed for each image,\n        resulting in 30 features.  For instance, field 0 is Mean Radius, field\n        10 is Radius SE, field 20 is Worst Radius.\n\n        - class:\n                - WDBC-Malignant\n                - WDBC-Benign\n\n    :Summary Statistics:\n\n    ===================================== ====== ======\n                                           Min    Max\n    ===================================== ====== ======\n    radius (mean):                        6.981  28.11\n    texture (mean):                       9.71   39.28\n    perimeter (mean):                     43.79  188.5\n    area (mean):                          143.5  2501.0\n    smoothness (mean):                    0.053  0.163\n    compactness (mean):                   0.019  0.345\n    concavity (mean):                     0.0    0.427\n    concave points (mean):                0.0    0.201\n    symmetry (mean):                      0.106  0.304\n    fractal dimension (mean):             0.05   0.097\n    radius (standard error):              0.112  2.873\n    texture (standard error):             0.36   4.885\n    perimeter (standard error):           0.757  21.98\n    area (standard error):                6.802  542.2\n    smoothness (standard error):          0.002  0.031\n    compactness (standard error):         0.002  0.135\n    concavity (standard error):           0.0    0.396\n    concave points (standard error):      0.0    0.053\n    symmetry (standard error):            0.008  0.079\n    fractal dimension (standard error):   0.001  0.03\n    radius (worst):                       7.93   36.04\n    texture (worst):                      12.02  49.54\n    perimeter (worst):                    50.41  251.2\n    area (worst):                         185.2  4254.0\n    smoothness (worst):                   0.071  0.223\n    compactness (worst):                  0.027  1.058\n    concavity (worst):                    0.0    1.252\n    concave points (worst):               0.0    0.291\n    symmetry (worst):                     0.156  0.664\n    fractal dimension (worst):            0.055  0.208\n    ===================================== ====== ======\n\n    :Missing Attribute Values: None\n\n    :Class Distribution: 212 - Malignant, 357 - Benign\n\n    :Creator:  Dr. William H. Wolberg, W. Nick Street, Olvi L. Mangasarian\n\n    :Donor: Nick Street\n\n    :Date: November, 1995\n\nThis is a copy of UCI ML Breast Cancer Wisconsin (Diagnostic) datasets.\nhttps://goo.gl/U2Uwz2\n\nFeatures are computed from a digitized image of a fine needle\naspirate (FNA) of a breast mass.  They describe\ncharacteristics of the cell nuclei present in the image.\n\nSeparating plane described above was obtained using\nMultisurface Method-Tree (MSM-T) [K. P. Bennett, "Decision Tree\nConstruction Via Linear Programming." Proceedings of the 4th\nMidwest Artificial Intelligence and Cognitive Science Society,\npp. 97-101, 1992], a classification method which uses linear\nprogramming to construct a decision tree.  Relevant features\nwere selected using an exhaustive search in the space of 1-4\nfeatures and 1-3 separating planes.\n\nThe actual linear program used to obtain the separating plane\nin the 3-dimensional space is that described in:\n[K. P. Bennett and O. L. Mangasarian: "Robust Linear\nProgramming Discrimination of Two Linearly Inseparable Sets",\nOptimization Methods and Software 1, 1992, 23-34].\n\nThis database is also available through the UW CS ftp server:\n\nftp ftp.cs.wisc.edu\ncd math-prog/cpo-dataset/machine-learn/WDBC/\n\n.. topic:: References\n\n   - W.N. Street, W.H. Wolberg and O.L. Mangasarian. Nuclear feature extraction \n     for breast tumor diagnosis. IS&T/SPIE 1993 International Symposium on \n     Electronic Imaging: Science and Technology, volume 1905, pages 861-870,\n     San Jose, CA, 1993.\n   - O.L. Mangasarian, W.N. Street and W.H. Wolberg. Breast cancer diagnosis and \n     prognosis via linear programming. Operations Research, 43(4), pages 570-577, \n     July-August 1995.\n   - W.H. Wolberg, W.N. Street, and O.L. Mangasarian. Machine learning techniques\n     to diagnose breast cancer from fine-needle aspirates. Cancer Letters 77 (1994) \n     163-171.',
 'feature_names': array(['mean radius', 'mean texture', 'mean perimeter', 'mean area',
        'mean smoothness', 'mean compactness', 'mean concavity',
        'mean concave points', 'mean symmetry', 'mean fractal dimension',
        'radius error', 'texture error', 'perimeter error', 'area error',
        'smoothness error', 'compactness error', 'concavity error',
        'concave points error', 'symmetry error',
        'fractal dimension error', 'worst radius', 'worst texture',
        'worst perimeter', 'worst area', 'worst smoothness',
        'worst compactness', 'worst concavity', 'worst concave points',
        'worst symmetry', 'worst fractal dimension'], dtype='<U23'),
 'filename': 'breast_cancer.csv',
 'data_module': 'sklearn.datasets.data'}

划分训练集与测试集。

x_train, x_test, y_train, y_test = train_test_split(
    cancers.data, cancers.target, test_size=0.20)
print("x_train.shape:", x_train.shape)
print("y_train.shape:", y_train.shape)
print("x_test.shape:", x_test.shape)
print("y_test.shape:", y_test.shape)
# 标准化数据,保证每个维度的特征数据方差为1,均值为0,使得预测结果不会被某些维度过大的特征值而主导
ss = StandardScaler()
# fit_transform()先拟合数据,再标准化
x_train = ss.fit_transform(x_train)
# transform()数据标准化
x_test = ss.transform(x_test)
x_train.shape: (455, 30)
y_train.shape: (455,)
x_test.shape: (114, 30)
y_test.shape: (114,)

建立BP模型(评估后可进行调参,从而选择最优参数)。

from sklearn.neural_network import MLPClassifier
# 建立 BP 模型, 采用Adam优化器,relu非线性映射函数
BP = MLPClassifier(solver='adam',activation = 'relu',max_iter = 1000,alpha = 1e-3,hidden_layer_sizes = (64,32, 32),random_state = 1)

进行模型训练。

BP.fit(x_train, y_train)
MLPClassifier(alpha=0.001, hidden_layer_sizes=(64, 32, 32), max_iter=1000,
              random_state=1)

进行模型预测,对真实数据和预测数据进行可视化(用Axes3D绘制3d散点图)。

predict_train = BP.predict(x_train)
#打印模型分数(预测精度)
print('测试数据集得分:{:.2f}%'.format(mlp_hw.score(x_test,y_test)*100))
print()
fig = plt.figure()
ax = Axes3D(fig, rect=[0, 0, 1, 1], elev=20, azim=20) 
ax.scatter(x_train[:, 0], x_train[:, 1], x_train[:, 2], marker='o', c=y_train)
plt.title('True Label Map')
plt.show()
# 可视化预测数据
fig = plt.figure()
ax = Axes3D(fig, rect=[0, 0, 1, 1], elev=20, azim=20) 
ax.scatter(x_train[:, 0], x_train[:, 1], x_train[:, 2], marker='o', c=predict_train)
plt.title('Cancer with BP Model')
plt.show()
测试数据集得分:99.12%

C:\Users\28599\AppData\Local\Temp\ipykernel_21744\3167179033.py:6: MatplotlibDeprecationWarning: Axes3D(fig) adding itself to the figure is deprecated since 3.4. Pass the keyword argument auto_add_to_figure=False and use fig.add_axes(ax) to suppress this warning. The default value of auto_add_to_figure will change to False in mpl3.5 and True values will no longer work in 3.6.  This is consistent with other Axes classes.
  ax = Axes3D(fig, rect=[0, 0, 1, 1], elev=20, azim=20)

apaang

C:\Users\28599\AppData\Local\Temp\ipykernel_21744\3167179033.py:12: MatplotlibDeprecationWarning: Axes3D(fig) adding itself to the figure is deprecated since 3.4. Pass the keyword argument auto_add_to_figure=False and use fig.add_axes(ax) to suppress this warning. The default value of auto_add_to_figure will change to False in mpl3.5 and True values will no longer work in 3.6.  This is consistent with other Axes classes.
  ax = Axes3D(fig, rect=[0, 0, 1, 1], elev=20, azim=20)

paang

进行模型评估,并进行预测结果指标统计(统计每一类别的预测准确率、召回率、F1分数)。

from sklearn.metrics import classification_report, confusion_matrix
# 显示预测分数
print("预测准确率: {:.4f}".format(BP.score(x_test, y_test)))
# 进行测试集数据的类别预测
predict_test = BP.predict(x_test)
print("测试集的真实标签:\n", y_test)
print("测试集的预测标签:\n", predict_test)
print(classification_report(y_test, predict_test))

预测准确率: 0.9825
测试集的真实标签:
 [0 1 1 0 1 1 0 0 1 0 0 1 1 0 1 1 1 1 1 0 0 1 1 0 0 1 0 1 1 0 0 1 1 0 1 1 1
 0 0 0 1 0 0 1 1 1 1 1 1 0 1 0 0 1 1 1 0 0 1 0 1 1 1 0 0 1 0 1 0 0 1 0 0 1
 0 1 1 0 1 1 1 1 0 0 1 1 0 0 1 1 1 1 0 1 0 0 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1
 1 0 1]
测试集的预测标签:
 [0 1 1 0 1 1 0 0 1 0 0 1 1 0 1 1 1 1 1 0 0 1 1 0 0 1 0 1 1 0 0 1 1 0 1 1 1
 0 1 0 1 0 0 1 1 1 1 1 1 0 1 0 0 1 1 1 0 1 1 0 1 1 1 0 0 1 0 1 0 0 1 0 0 1
 0 1 1 0 1 1 1 1 0 0 1 1 0 0 1 1 1 1 0 1 0 0 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1
 1 0 1]
              precision    recall  f1-score   support

           0       1.00      0.96      0.98        45
           1       0.97      1.00      0.99        69

    accuracy                           0.98       114
   macro avg       0.99      0.98      0.98       114
weighted avg       0.98      0.98      0.98       114

计算混淆矩阵,并用热力图显示。

#打印混淆矩阵
from sklearn.metrics import confusion_matrix
# 计算混淆矩阵
confusion_mat = confusion_matrix(y_test, predict_test)
# 打混淆矩阵
print('混淆矩阵:')
print(confusion_mat)

# 将混淆矩阵以热力图显示
import seaborn as sns
sns.set()
figure, ax = plt.subplots()
# 画热力图
sns.heatmap(confusion_mat, cmap="YlGnBu_r", annot=True, ax=ax)  
# 标题 
ax.set_title('confusion matrix')
# x轴为预测类别
ax.set_xlabel('predict')  
# y轴实际类别
ax.set_ylabel('true')  
plt.show()


混淆矩阵:
[[43  2]
 [ 0 69]]

pang

posted @   嘿,抬头!  阅读(292)  评论(0编辑  收藏  举报
相关博文:
阅读排行:
· 全程不用写代码,我用AI程序员写了一个飞机大战
· DeepSeek 开源周回顾「GitHub 热点速览」
· 记一次.NET内存居高不下排查解决与启示
· 物流快递公司核心技术能力-地址解析分单基础技术分享
· .NET10 - 预览版1新功能体验(一)
点击右上角即可分享
微信分享提示