Thread by @jslbutler

Here is my full playlist for @neuromatch Deep Learning.
My song for Basics And Pytorch (W1D1) by Shubh Pachchigar, @haltakov , Matthew Sargent, @KordingLab is
“Rolling in the Deep” by Adele
1/16

My song choice for Linear Deep Learning (W1D2) by Saeed Salehi, @haltakov & @SaxeLab is
River Deep Mountain High by Ike & Tina Turner
because
"And the ML gets stronger in every way
And it gets deeper, deeper, deeper, let me say" 2/16

My song choice for Multi Layer Perceptrons (W1D3) by @Arash2ML, @SuryaGanguli is
"How deep is your Love" by Take That
because we need to know,
"How deep is your Layer"

https://twitter.com/jslbutler/status/1422093095156453377?s=20

3/16

My song choice for Optimization (W1D4) by @JGallegoPosada @bouzoukipunks is
"In too Deep" by Sum 41
"'Cause I'm in too deep, and I'm trying to keep
Up above in my head, instead of going under"
Instead of the going under students learn DL @neuromatch 4/16

My song choice for Regularization (W1D5) by Ravi Teja Konkimalla, Mohitrajhu Lingan Kumaraian, Kevin Machado Gamboa, @Kelson_SS & @LyleUngar is
"Regulate" by Warren G & Nate Dogg 5/16

My song choice for Convnets And RNNs (W2D1) by Dawn Estes McKnight, Richard Gerum, Cassidy Pirlot, Rohan Saha, Liam Peet-Pare, Saeed Najafi, @alonamarie is
"Share Your Love with Me" by Aretha Franklin,
As Convnets and RNNs should share your parameters 6/16

My song choice for Modern Convnets (W2D2) by Laura Pede, Richard Vogg, Marissa Weis, Timo Lüddecke, @alxecker is
"Modern Love" by David Bowie
As you would not have a problem recognising David Bowie's face with a Covent
7/16

My song choice for Modern Recurrent Neural Networks (W2D3) by @bhargavvader, @AZahedifard, @profjamesevans is
"The Modern Age" by The Strokes
As I'm not sure if you would need an RNN to predict the next note of the song but it is still great 8/16

My song choice for Attention And Transformers (W2D4) by
Bikram Khastgir, @RajaswaPatil, Egor Zverev, He He is
"Walk on the Wild Side" by Lou Reed from the album Transformer
As to Break the model you have to talk on the wild side (see tutorial) 9/16

My song choice for Generative Models (W2D5) by Saeed Salehi, Spiros Chavlis, Vikash Gilja is
"Generator" by The Holloways
Because you can use a GAN to generate the music that makes you feel better.
10/16

My song choice for Unsupervised And Self Supervised Learning (W3D1) by Arna Ghosh, Colleen Gillon, Tim Lillicrap, @tyrell_turing is
"everybody's gotta learn sometime" by the Korgis
Because every algorithms got to learn sometime
11/16

My song choice for Basic Reinforcement Learning (W3D2) by Matthew Sargent, Anoop Kulkarni, Sowmya Parthiban, @FeryalMP, @janexwang is
"You Learn" by Alanis Morissette
because the song describes the concepts of reinforcement learning 12/16

My song choice for Reinforcement Learning For Games (W3D3) by Mandana Samiei, Raymond Chua, Tim Lilicrap, @tyrell_turing is
Games by Tessa Violet
As you have to keep playing games to learn (see song lyrics and tutorial) 13/16

My song choice for Continual Learning (W3D4) by @kw_cooper, Diganta Misra, Gido van de Ven, Andrea Cossu, @v_lomonaco is
"Still Learning " by Halsey 14/16

Here are all my suggested @neuromatch deep learning songs themed playlist 15/16
open.spotify.com/playlist/4oixP…

@neuromatch And most important of all the @neuromatch deep learning materials that inspired the playlist, once again expert colour advice provided my daughter.
deeplearning.neuromatch.io/tutorials/intr… 16/16

@threadreaderapp unroll