song reccomendation using hybrid models

Inspiration

Inspired by the recommendation systems used by Spotify, Apple Music, and YouTube, we aimed to build a hybrid recommendation model that combines Collaborative Filtering (CF) and Content-Based Filtering (CBF) to provide accurate and personalized song recommendations.

What it does

By combining CF and CBF, we ensure more diverse and relevant song recommendations.

Challenges we ran into

hybrid model tuning, balancing CF and CBF contributions to ensure relevant recommendations performance optimization, optimizing Annoy search parameters for fast recommendations.

Accomplishments that we're proud of

successfully implemented a hybrid recommendation model, achieved meaningful, personalized recommendations for users.

What we learned

how to optimize nearest-neighbor search using Annoy, the importance of feature scaling and matrix factorization.

What's next for song reccomendation using hybrid models

develop strategies to recommend songs for new users with no history.

Built With

• annoy
• css
• html
• javascript
• numpy
• pandas
• python
• scikit-learn
• scipy