Akshay-Rajesh / Spam-text-classifier

Repository for mini projects
0 stars 1 forks source link

Model is overfitting , explore more ways to clean ? #19

Closed Akshay-Rajesh closed 2 days ago

pffaundez commented 6 days ago

Instead of an overfitting, it shows a probable effect of inbalanced data since it seems to be biasing the model to always predict the HAM class instead of the minority class (SPAM). Something observable from the 1.0 obtained in Precision. Let's discuss in our next weekly about balancing methods for text data.