# Rafael Almeida — ISM4542 Statistical Programming for Business Analytics
# Module 8: logistic regression for spam classification
library(rio)
emails <- import("4542 Module 8 Assignment Data.xlsx")
set.seed(54851367)
sample_data <- emails[sample(seq_len(nrow(emails)), 1000), ]
str(sample_data)
spam_model <- glm(spam ~ . - row, family = binomial, data = sample_data)
summary(spam_model)
confint(spam_model)
predicted_class <- ifelse(spam_model$fitted.values >= 0.5, 1, 0)
confusion_matrix <- table(Predicted = predicted_class, Actual = sample_data$spam)
confusion_matrix
accuracy <- mean(predicted_class == sample_data$spam)
accuracy
table(sample_data$urgent_subj)