什么是逻辑回归
应用场景
逻辑回归的原理
掌握逻辑回归,必须掌握以下两点
- 逻辑回归中,其输入值是什么
- 如何判断逻辑回归的输出
输入
激活函数
衡量损失
损失
优化
API
肿瘤预测案例
数据介绍
代码实现
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
# 获取数据
names = ['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=names)
data.head()
# 数据基本处理
# 缺失值处理
data = data.replace(to_replace="?",value=np.nan)
data = data.dropna()
# 确定特征值,目标值
x = data.iloc[:,1:-1]
y = data["Class"]
# 分割数据
x_train,x_test,y_train,y_test = train_test_split(x,y,random_state=22,test_size=0.2)
# 特征工程 标准化
transfer = StandardScaler()
x_train = transfer.fit_transform(x_train)
x_test = transfer.transform(x_test)
# 机器学习
estmator = LogisticRegression()
estmator.fit(x_train,y_train)
# 模型评估
print("准确率:\n",estmator.score(x_test,y_test))
print("预测值:\n",estmator.predict(x_test))
分类评估方法
准确率与召回率
混淆矩阵
我们之前使用的准确率公式为:(TP+TN)/(TP+Fn+FP+TN)
精确率(Precision)与召回率(Recall)
精确率:(TP)/(TP+FP)
召回率:(TP)/(TP+FN)
F1-score
分类评估报告api
from sklearn.metrics import classification_report
y_pre = estmator.predict(x_test)
ret = classification_report(y_test,y_pre,labels=(2,4),target_names=("良性","恶性"))
print(ret)
ROC曲线与AUC指标
TPR与FPR
ROC曲线
AUC指标
AUC计算API
from sklearn.metrics import roc_auc_score
y_test = np.where(y_test>3,1,0)
roc_auc_score(y_test,y_pre)
解决类别不平衡问题
pip3 install imbalanced-learn
准备类别不平衡数据
from sklearn.datasets import make_classification
import matplotlib.pylab as plt
from collections import Counter
X,Y = make_classification(n_samples=5000,
n_features=2, # 特征个数= n_informative()+ n_redundant()+ n_repeated()
n_informative=2,# 多信息特征的个数
n_redundant=0,# 冗余信息,informative特征的随机线性组合
n_repeated=0,# 重复信息,随机提取n_informative和n_redundant特征
n_classes=3,# 分类类别
n_clusters_per_class=1,# 某一个类别是由几个cluster构成的
weights=[0.01,0.05,0.94],# 列表类型,权重比
random_state=0)
X,Y,X.shape
Counter(y)
# 数据可视化
plt.scatter(X[:,0],X[:,1],c=Y)
plt.show()
解决办法
过采样方法
随机过采样方法
from imblearn.over_sampling import RandomOverSampler
ros = RandomOverSampler(random_state=0)
X_resampled,Y_resampled = ros.fit_resample(X,Y)
Counter(Y_resampled)
# 数据可视化
plt.scatter(X_resampled[:,0],X_resampled[:,1],c=Y_resampled)
plt.show()
过采样代表性算法-SMOTE
from imblearn.over_sampling import SMOTE
X_resampled,Y_resampled = SMOTE().fit_resample(X,Y)
Counter(Y_resampled)
# 数据可视化
plt.scatter(X_resampled[:,0],X_resampled[:,1],c=Y_resampled)
plt.show()
欠采样方法
随机欠采样方法
from imblearn.under_sampling import RandomUnderSampler
rus = RandomUnderSampler(random_state=0)
X_resampled,Y_resampled = rus.fit_resample(X,Y)
Counter(Y_resampled)
# 数据可视化
plt.scatter(X_resampled[:,0],X_resampled[:,1],c=Y_resampled)
plt.show()