sklearn学习笔记（11）——分类算法-逻辑回归

逻辑回归

sigmoid函数：

逻辑回归公式：

 

逻辑回归损失函数：

与线性回归原理相同，但由于是分类问题，损失函数不一样，只能通过梯度下降求解

对数似然损失函数：

完整的损失函数：

注：cost损失的值越小，那么预测的类别准确度越高

 

sklearn逻辑回归API：sklearn.linear_model.LogisticRegression

LogisticRegression总结：

优点：适合需要得到一个分类概率的场景，简单，速度快

缺点：不好处理多分类问题

应用：广告点击率、是否患病、金融诈骗、是否为虚假账号

---

# -*- coding: utf-8 -*-

import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import classification_report


def my_logistic():
    """
    逻辑回归做二分类，癌症预测
    :return: None
    """
    # 读取数据，指定数据读取的列名
    column = ['Sample code number', 'Clump Thickness',
              'Uniformity of Cell Size', 'Uniformity of Cell Shape',
              'Marginal Adhesion', 'Single Epithelial Cell Size',
              'Bare Nuclei', 'Bland Chromatin', 'Normal Nucleoli',
              'Mitoses', 'Class']
    data = pd.read_csv(
        "https://archive.ics.uci.edu/ml/machine-learning-databases/breast-cancer-wisconsin/breast-cancer-wisconsin.data",
        names=column)
    # 替换?为np.nan, 删除?标记的缺失值
    data = data.replace(to_replace="?", value=np.NaN)
    data = data.dropna()
    print(data)

    # 进行数据分割，取出特征值和目标值
    x = data[column[1: 10]]
    y = data[column[10]]

    x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.25)

    # 进行数据的标准化
    std = StandardScaler()
    x_train = std.fit_transform(x_train)
    x_test = std.transform(x_test)

    # 逻辑回归预测
    lr = LogisticRegression()
    lr.fit(x_train, y_train)
    print("得出的权重参数为：", lr.coef_)

    y_predict = lr.predict(x_test)

    print("预测的精确率和召回率为：", "\n", classification_report(y_test, y_predict, 
                                                      labels=[2, 4], 
                                                      target_names=["良性", "恶性"]))


if __name__ == '__main__':
    my_logistic()