Skip to content

Latest commit

 

History

History
2 lines (2 loc) · 227 Bytes

File metadata and controls

2 lines (2 loc) · 227 Bytes

Identify-spam-in-social-listening-data

The project did pre-processing on tweets, used suspended account information to label tweets as spam, and finally built a logistic regression and a MLP model to predict an unseen tweet.