MACHINE LEARNING TECHNIQUES FOR MUSIC PREDICTION S. Grant Lowe
Advisor: Prof. Nick Webb
RESEARCH QUESTIONS
Can we predict the year in which a song was released?
Can we predict the genre of a song?
Can we identify which attributes are the strongest in answering these questions?
BACKGROUND
Hit Song Science Genre Classification Year Prediction
APPROACH
Use WEKA Use the Million Song Dataset
WEKA
Machine Learning Software Contains Visualization tools and
algorithms for data analysis and modeling
DATA
Million Song Data Set: commercial tracks from 1922-2011,collected by LabROSA
EARLY CHALLENGES
Data in the wrong Format: HDF5 vs CSV Lots of missing Data! Almost half of the songs are missing
year, a very important attribute Many attributes are being ignored
because a majority of the songs are missing data.
ArtistID -> Year?
ATTRIBUTES
The MSD contains 53 descriptive attributes for each song, along with 90 timbre attributes. Attributes were removed if they were not good indicators of release year or genre, or if they were too closely tied to what was being classified.
ATTRIBUTE MOTIVATION
Ranked Descriptive Attributes• Loudness (measured in decibels)• Duration (in seconds)• Tempo (estimated tempo in BPM)• Time Signature (estimated beats per bar)• Key • Mode (major or minor)
Timbre is the quality of a musical note or sound that distinguishes different types of musical instruments, or voices. It is a complex notion also referred to as sound color, texture, or tone quality, and is derived from the shape of a segment’s spectro-temporal surface, independently of pitch and loudness.
EARLY RESULTS – DESCRIPTIVE ATTRIBUTES Discretized into 6 decades; 1960-1970, 1970-1980, etc. Baseline (Chance selection): 16.67% First Tests: 6-9% correctly classified More recent Tests: 25-30% Why Random Forest and BayesNet?
EARLY RESULTS
TIMBRE RESULTS
GENRE PREDICTION
Genres: Classic pop and rock Classical Dance and Electronica Folk Hip-Hop Jazz Metal Pop Rock and Indie Soul and Reggae
GENRE PREDICTION RESULTS
CONCLUSIONS & FUTURE WORK
Timbre Attributes are better than Descriptive Attributes – Why?
Taste Profile Lyrical/Emotional Content Tag Dataset
QUESTIONS?