This project focused on the comparison of three different recommendation strategies to automate playlist continuation. The models were trained and evaluated on a massive dataset of 100K playlists containing over 600K songs

Experimental Approaches

• 01 Collaborative Filtering: Comparative analysis of user-based and item-based filtering techniques.
• 02 Neural Networks: Implementation of a deep learning approach for sequence-based song prediction.
• 03 Big Data Pipeline: Utilized PySpark for large-scale data handling and Petastorm for efficient PyTorch data loading.