Kaggle Competition: Avito Demand Challenge

This project was a group submission - in collaboration with Minxuan Chen, Parisa Mofakham, and Huilin Cai – for a Kaggle contest to predict demand for online advertisements. The ensemble model used image analysis, NLP, neural networks and gradient boosting, producing a score of 0.22402 (winning score was 0.21501; lower scores are better).

Full code and documentation can be found here.