Music Recommendation Service

Using machine learning to predict if a given song will be popular or not

Project for CMU 15-780 Graduate Artificial Intelligence

Mining Million Song Dataset

Processing the Million Song Dataset with Apache Spark

Predict whether a song is 'hot' or not, through analysis of the Million Song Dataset.

Design and implementation from scratch of different models for a musical recommendation system

Notebooks and data to accompany Python instruction for Data in Social Context Fall 2018

Discovery Recommender System

Data Mining course project - Million Songs Dataset exploration

Recommendation system on Million Song Dataset

Classification of audio features using different ML algorithms on the MSD. The project was done for Machine Learning module at Coventry University.

Tools to run text-based experiments for large scale cover detection.

An R project that investigates whether different genres of songs have significantly different durations through the use of a one-way ANOVA test and post hoc significance tests conducted over an excerpt of a dataset consisting of 1 million popular songs compiled by The Echo Nest and a lab at Columbia University.

Trend analysis of pop music and prediction of release year.

An online song recommender based on a K-means model using the Spotify API and the MillionSongSubset

Classying music genre based on audio features and lyrics (using the Million Song Dataset).

Language clustering of the musicXmatch dictionary in the Million Song Dataset