About Shopee codeleague
Shopee Code League is a 2-month online coding challenge consisting of a series of competitions, online algorithm questions and online training workshops open to all students and professionals across the region. The competitions are specially designed by the Shopee tech teams. Competition link.
Our work
| Task | Description | Ranking | Notebook |
|---|---|---|---|
| Product Detection | Developed an ensemble model of Resnet50, Resnet152, Inception V3 and Inception Resnet V2 to classify various product categories (42 categories) based on product images posted by seller in Shopee platform. | 63rd out of 823 teams (top 8%) | [Link] |
| Logistics (Data Analytics) | Identified all the orders that are considered late depending on the Service Level Agreements (SLA). | 96th out of 547 teams (top 18%) | [Link] |
| Sentiment Analysis | Built an ensemble model of GPT-2, Roberta and BERT for predicting review score (1-5 stars) from user product review. | 2nd out of 356 teams (top 1%) | [Link] |
| Marketing Analytics (Problem like fruad detection) | Applied resampling techniques, Tomek Link and SMOTE, to reduce imbalace of labels and built an tree-based model, Catboost, to predict whether a user opens the emails sent by Shopee. | 24th out of 368 teams (top 7%) | [Link] |
| Fake transaction detection (Practice problem) | Detected sellers who perform fake transactions to boost their sales/shop ratings using dictionary, Python, for efficient searching. | - | [Link] |
- © Untitled
- Design: HTML5 UP