[M]-Confusion-Matrix.md

Confusion Matrix

Preparation Code

# Sample Data
library(dplyr)
M_train <- mtcars %>% filter(mpg <= 24)
M_test <- mtcars %>% filter(mpg > 24)
fit <- glm(am ~ mpg + cyl, family = "binomial", M_train) # `am` is binary, `mpg` and `cyl` are continuous

Actual Code

M_test$predicted_probability <- predict(fit, M_test, type = "response")

M_test$predicted <- factor(round(M_test$predicted_probability),
                           levels = c(0,1),
                           labels = c("0_predict","1_predict"))

table(M_test$am, M_test$predicted)