Movie Recommendation System
Movie recommendation: from Matrix Factorization to hybrid Neural CF, benchmarked against a zero-shot LLM.
ML Developer (with Louis Rollet & Vincent Montero-Fontaine)
A complete movie recommendation system on the MovieLens 20M dataset (PyTorch): Matrix Factorization baseline (RMSE ~1.16), Neural Collaborative Filtering (RMSE ~0.82) and a hybrid extension integrating genre embeddings to handle cold start, all benchmarked against a zero-shot LLM approach (Gemini, RMSE ~0.99). Deployed as an interactive Streamlit dashboard: profile-based recommendations, similar-movie search via cosine similarity on learned embeddings, and a taste-profile builder for new users through a "proxy user" algorithm.