Classifies Spam and Ham emails using Naive Bayes and Logistic Regression
This Repo contains 3 Python scripts, Logistic Regression with and without stop words. The stop words like "I, The, am etc." are removed.
Without removing the stop words the Naive Bayes produced an accuracy of 92.47 %, whislt with removing the stop words the accuracy dropped down to 90.38%.
In Logistic Regression, the iteration was set to 50 with regularization parameter and learning rate set to 0.01. It yielded 92.25 % accuracy.