逻辑回归模型

R 统计面试题实战练习

Zuzanna Chmielewska

Actuary

逻辑回归的应用

垃圾邮件

金融欺诈

二元数据点

$$f(x) = \frac{1}{1+e^{-x}} $$

$$f(x) = \frac{1}{1+e^{-x}} \in (0,1)$$

概率预测： $$ p_i = P(y_i=1) = \frac{1}{1+e^{-(\beta_0+\beta_1 \cdot x_{i1}+...+\beta_p \cdot x_{ip})}} $$

对数几率预测： $$ l_i = ln(\frac{p_i}{1-p_i}) = \beta_0+\beta_1 \cdot x_{i1}+...+\beta_p \cdot x_{ip}$$

二元数据点的预测

model <- glm(y ~ x, data = df, family = "binomial")

model <- glm(y ~ x, data = df, family = "binomial")
predict(model, newdata = new_df, type = "response")

R 统计面试题实战练习