Skip to content

Latest commit

 

History

History
24 lines (13 loc) · 789 Bytes

README.md

File metadata and controls

24 lines (13 loc) · 789 Bytes

Click Through Rate Prediction

Coursework Project, Learning with Imbalanced Data

ML program that can predict whether someone will click on an online ad as it crucial for understanding how well ads are doing, which is important for search engines.

Model Architecture

Model Architecture

Dataset

Avazu Ads Dataset

Class Imbalance : There are less number of clicks on ads compared to non-clicks

Attributes : click: non-click/click , site_id , site_domain , site_category , app_id , app_domain , app_category , device_id , device_ip , device_model , device_type , device_conn_type.

Imbalance Learning Strategies Used

  • SMOTE
  • Oversampling
  • Undersampling