Oragnised by Consulting and Analytics club, IIT Guwahati
Link to competition [Datathon]
In this world of big data, Trell wants you to use the data to predict the age group of their users based on their activity on social media activities. This will help them to divide their huge userbase and cater differently to each of them. Given this huge dataset, predict the age group of the users, the evaluation metric for the competition is the Weighted F1 score.
Train dataset [link]
Test dataset [link]
The model folder consists of the final model (Team_Ragrun_cascade_cup.ipynb) I chose for submission, along with the corresponding prediction csv file (prediction.csv). It gave a f1 score of 80.988 on public leaderboard and 81.14 on private leaderboard