Reproduce “SpotHitPy: A Study For ML-Based Song Hit Prediction Using Spotify” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Ioannis Dimolitsas, Spyridon Kantarelis, Afroditi Fouka
Researchers at the National Technical University of Athens developed SpotHitPy, a system leveraging machine learning to predict hit songs using Spotify audio features and Billboard chart data from 2011-2021. Their Random Forest model achieved approximately 86% accuracy in distinguishing hits from non-hits after balancing the dataset and applying dimensionality reduction.
Key insights
- Audio features from Spotify contain quantifiable patterns that machine learning models can use to differentiate between hit and non-hit songs.
- Addressing data imbalance and performing dimensionality reduction (PCA) are crucial steps for effective hit song prediction.
- Ensemble methods like Random Forest are effective in capturing complex, non-linear relationships present in musical audio features.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.