Just spent a day at MLConf where I was talking about how we do music recommendations. There was a whole range of great speakers (actually almost 2/3 women which was pretty cool in itself). Here are my slides:
From: Erik Bernhardsson
Scrolling through the Discover page on Spotify the other day it occurred to me that the album is in fact a fairly strong visual proxy for what kind of content you can expect from it. I started wondering if the album cover can in fact be used for recommendations.
From: Erik Bernhardsson
In case you missed it, we just acquired a company called Echo Nest in Boston. These people have been obsessed with understanding music for the past 8 years since it was founded by Brian Whitman and Tristan Jehan out of MIT Medialab.
From: Erik Bernhardsson
So Luigi, our open sourced workflow engine in Python, just recently passed 1,000 stars on Github, then shortly after passed mrjob as (I think) the most popular Python package to do Hadoop stuff. This is exciting!
From: Erik Bernhardsson
Haven't posted anything in ages, so here's a quick hack I threw together in Python on a Sunday night.
From: Erik Bernhardsson
We run a ton of A/B tests at Spotify and we look at a ton of metrics. Defining metrics is a little bit of an art form. Ideally you want to define success metrics before you run a test to avoid cherry picking metrics.
From: Erik Bernhardsson
Radim Rehurek has put together an excellent summary of approximate nearest neighbor libraries in Python. This is exciting, because one of the libraries he's covering, annoy, was built by me. After introducing the problem, he goes through the list of contestants and sticks with five remaining ones.
From: Erik Bernhardsson
I wanted to share some more insight into the algorithms we use at Spotify. One matrix factorization algorithm we have used for a while assumes that we have user vectors $$ bf{a}_u $$ and item vectors $$ bf{b}_i $$ .
From: Erik Bernhardsson
I think it's funny how MS at some point realized they are not the cool kids and there's no reason to appeal to that target audience. Their new marketing strategy finally admits what's been long known: the correlation between “business casual” and using Microsoft products:
From: Erik Bernhardsson
One thing I encountered today was a trick using bagging as a way to go beyond a point estimate and get an approximation for the full distribution. This can then be used to penalize predictions with larger uncertainty, which helps reducing false positives.
From: Erik Bernhardsson
A lot of people have asked me what models we use for recommendations at Spotify so I wanted to share some insights. Here's benchmarks for some models. Note that we don't use all of them in production.
From: Erik Bernhardsson
Btw I just put something up online that I spent a couple of evenings in my couch putting together: it's a website where you can track any numerical data on the web. Want to know how many Twitter followers you have?
From: Erik Bernhardsson
A lot of people these days know about collaborative filtering.
From: Erik Bernhardsson
If you have a few minutes, you should check out mine and Chris Johnsonâs panel proposal.
From: Erik Bernhardsson
I just answered a Quora question about what, if any, are the differences in the algorithms that are behind recommendations for music and movies.
From: Erik Bernhardsson
Andy Sloane decided to call my 2D visualization and raise it to 3D. (Looks a little weird in the iframe but check out the link). It's based on a LDA model with 200 topics, so the artists tend to stick to clusters where each cluster is a topic.
From: Erik Bernhardsson
I've turned into a lazy bastard and I'm just posting presentations on this blog, but here's one from Rohan Singh at Spotify talking about the backend infrastructure of the Discover page.
From: Erik Bernhardsson
I was just at the NYC Predictive Analytics meetup talking about how we build machine learning algorithms using Hadoop to power music recommendations. Great meetup, where we had two speakers, me and Blake Shaw from Foursquare.
From: Erik Bernhardsson
I thought this article about the company culture at HubSpot is kind of funny. “HubSpot's Awesome Presentation Shows how to Create a 21st Century Culture”. Just FYI: You're not different. You're a bunch of white hipsters aged 25-30 dressed up in the same theme.
From: Erik Bernhardsson
