addressalign-toparrow-leftarrow-rightbackbellblockcalendarcameraccwcheckchevron-downchevron-leftchevron-rightchevron-small-downchevron-small-leftchevron-small-rightchevron-small-upchevron-upcircle-with-checkcircle-with-crosscircle-with-pluscrossdots-three-verticaleditemptyheartexporteye-with-lineeyefacebookfolderfullheartglobegmailgooglegroupshelp-with-circleimageimagesinstagramlinklocation-pinm-swarmSearchmailmessagesminusmoremuplabelShape 3 + Rectangle 1ShapeoutlookpersonJoin Group on CardStartprice-ribbonShapeShapeShapeShapeImported LayersImported LayersImported Layersshieldstartickettrashtriangle-downtriangle-uptwitteruserwarningyahoo

Deep Learning for Music Information Retrieval & Kaggle Galaxy Zoo Challenge

Excited to announce the May meetup where Sander will talk about his research and his winning entry for Kaggle's Galazy Zoo Challenge.

See you there!

Dirk & Ali

• Start: 6:30PM

• 6:30-45: Community updates and upcoming courses

• 6:45-8:15: Talk and Q&A

• End: 8:15 PM


UCL (University College London)

Christopher Ingold Building

XLG2 Auditorium

20 Gordon Street

WC1H 0AJ, London

Deep Learning for Music Information Retrieval (MIR) & Kaggle Galaxy Zoo Challenge

Deep learning has become a very popular approach for solving speech recognition and computer vision problems in recent years. In this talk we'll explore two different, but related applications. One is feature learning for music information retrieval (MIR): how can we use deep learning techniques to learn features from musical audio signals that are useful for classification and recommendation? We'll look at a few different tasks and feature learning approaches.

The other is galaxy morphology prediction: by automatically classifying galaxies based on their shape, astronomers can come to new insights about their origin and their distribution in space. We'll take a closer look at the convolutional neural network that won the recently finished Galaxy Zoo Challenge on Kaggle.

Bio: Sander Dieleman is a PhD student in the Reservoir Lab of prof. Benjamin Schrauwen at Ghent University in Belgium. His main research focus is applying deep learning and feature learning techniques to music information retrieval (MIR) problems, such as audio-based music classification, automatic tagging and music recommendation.

Thanks to UCL for providing the venue and our sponsors Learning Connexions.

Join or login to comment.

  • Rishabh M.

    Hi all,
    AFAIRemember, Sander mentioned about his Spotify internship during his talk in May. I just came across this post (**Recommending music on Spotify with deep learning**) wherein he describes his initial approach towards music recommendation while interning with Spotify; hope you find it useful. Thanks Sander!


    2 · August 6, 2014

  • dalton w.

    Hi I would like some quick tips on how to train to train a music dataset, the million song dataset subset using the deep learning toolbox in matlab/python. Thanks.

    June 18, 2014

  • Sander D.

    Hi everyone, thank you very much for the kind words! It was a lot of fun to present my work, and a new experience for me :)

    I've posted the slides online:

    Don't hesitate to get in touch if you have any more questions!


    7 · May 29, 2014

    • Christina N.

      Hi, congrats for the very nice talk! I do have a question about the recommendation factors, where the play counts where a strong positive (even when equal 1?) and play count = 0 was a weak negative. Perhaps a strong negative could be if the song was abandoned after a few seconds and then never returned to? that would be a clear indication of dis-likeness.

      June 9, 2014

    • Sander D.

      Yes, it would be :) Unfortunately the data that was available to us (from the MSD Taste Profile Subset) consisted of play counts, and I'm not sure how they were measuered, i.e. what constitutes a 'play'.

      June 9, 2014

  • Ali S.

    May 29, 2014

  • André B.

    Impressive talk, thanks! Do you plan to upload the slides somewhere?

    1 · May 29, 2014

    • Ali S.

      Yes, slides will be shared soon.

      1 · May 29, 2014

  • Matt

    Excellent presentation, very interesting, thanks!

    2 · May 29, 2014

  • João

    Thanks for your time, Sander.

    1 · May 29, 2014

  • Eve L.

    Very impressive work. Nice presentation. Thanks.

    3 · May 29, 2014

  • Amit T.

    My first deep learning meetup.. very interesting !!!

    1 · May 29, 2014

  • Jeff A.

    excellent presentation

    2 · May 29, 2014

  • Ali S.

    Best talk of 2014 on convnets.

    May 29, 2014

  • Ali S.

    Thank you Sander for an outstanding talk. We are very grateful you took the time out of your busy schedule. On behalf of the Deep Learning Meetup team, thank you for a great presentation.

    1 · May 29, 2014

  • Narmada G.

    Very informative talk, really enjoyed it...

    1 · May 28, 2014

  • Eloy

    thank you, really liked it Sander!

    1 · May 28, 2014

  • Nicolas C.

    Amazing pres - Thanks

    1 · May 28, 2014

  • Rene G.

    My first deep learning meetup: Very interesting and informative talk. Thanks. Are the slides available?

    3 · May 28, 2014

  • john d.

    Excellent talk, and a lot of food for thought
    Thank you

    2 · May 28, 2014

  • Peter N.

    Are there plans for a pub or similar afterwards? I'm excited by talks which mix astronomy and deep learning: I worked at Apache Point Observatory on the Sloan Digital Sky Survey, the imaging in the Galaxy Zoo, and now lead a deep learning startup. Double love!

    1 · May 28, 2014

  • Peter M.

    Note the new location!

    May 27, 2014

  • Kristoffer J.

    Looking forward

    May 26, 2014

  • Lionel W.

    I'm getting back to the UK a day too late for this... hope to join the next one!

    May 25, 2014

  • Peter M.

    Very interesting.

    May 19, 2014

Our Sponsors


    Meetup Organizer and Community Management


    Meetup Organizer and Funding

People in this
Meetup are also in:

Sign up

Meetup members, Log in

By clicking "Sign up" or "Sign up using Facebook", you confirm that you accept our Terms of Service & Privacy Policy