Machine Learning Meetings and Events
Group Meetings: Group meetings are held Mondays from 11am to Noon (talk starts 11:10am) in D.L. Pratt 290C unless otherwise noted. Meetings are coordinated by Hugo Larochelle.
Tea Talks: Tea talks are held every Wednesday at 4:00pm in D.L. Pratt 290C. Talks should be simple, accessible, and not exceed 15 minutes. Speakers bring snacks, make tea, and provide a copy of the presented paper.
Group Meeting Mar 2, 2009: Machine Learning for Music Genre Classification
- Speaker: James Bergstra (Montreal)
- Abstract:
What does machine learning have to do with music recommendation? With large online data collections and recommendation services, such as iTunes and last.fm, collaborative filtering and content-based recommendations become viable ways to connect artists with listeners. In this talk I’ll explain how machine learning is helping to develop computer understanding of musical taste. The first part will discuss music genre. I’ll argue that folksonomies from large communities of users, such as FreeDB and last.fm, are highly predictive of traditional notions of genre and are sufficiently fine-grained to characterize listeners’ tastes. These folksonomies provide semantic labels for the axes of a space in which geometric distance is related to stylistic similarity. The second part of the talk will describe machine learning methods for embedding songs and artists into this space by analysing the waveforms of music recordings. This is done by extracting features and then applying classification and regression algorithms.