Question

我有一个逻辑回归模型，其中所有变量都是虚拟的 (0-1)

如何将模型的系数转换为 SCORES？（我需要分数）

（经典方法都包括 binning - woebin 技术，但我这里不做 binning，因为我已经有了虚拟变量）

DF <- data.frame(
  Column1 = c(0,0,0,0,0,1,1,1,1,1), 
  Column2 = c(1,1,1,0,0,0,1,1,1,0),
  Column3 = c(1,1,1,0,0,0,0,0,0,0),
  TARGET = c(1,1,1,0,0,0,0,0,1,0)
)

log_model = glm(TARGET~. , family = "binomial", data = DF)

谢谢！

Answer 1

查看 here 并尝试将其应用于您的数据，也许它会有所帮助：

DF <- data.frame(
  Column1 = c(0,0,0,0,0,1,1,1,1,1), 
  Column2 = c(1,1,1,0,0,0,1,1,1,0),
  Column3 = c(1,1,1,0,0,0,0,0,0,0),
  TARGET = c(1,1,1,0,0,0,0,0,1,0)
)

log_model = glm(TARGET~. , family = "binomial", data = DF)

library(scorecardModelUtils)
DF$Y <- sample(0:1,size=nrow(DF),replace=TRUE)
x <- c("Column1","Column2","Column3")
iv_table_list <- iv_table(base = DF,target = "Y",num_var_name = x, cat_var_name = "TARGET")
num_cat <- num_to_cat(base = DF,num_woe_table = iv_table_list$num_woe_table)
scaling_tab <- scalling(base = num_cat,target = "TARGET",model = log_model)

虚拟变量逻辑回归到分数

1 个答案: