"What is the problem you are trying to solve? (And do I care about it?)". "Making sense" of a time series is a not a specific problem statement. I have watched for 15 minutes, and I wasted my time. This is mindless BS.

Brilliant interview. Thank you , David!

👏🏻

thank you! new ideas indeed!

Stamped with your presentation skill. All clear and lucid and 100 slides. 🎉❤🎉

Great Presentation... RIP TD Ameritrade

I've never thought about "no free lunch" the way he explained it, that there's always a price to pay in the form of a tradeoff, when you gain performance here you're likely loosing something else there. I always looked at it from the perspective that you can't just blindly apply a model on a dataset and expect the model to do all the work for you without any more input from you.

The legend

Great presentation. Need to check those tips about SORT BY are in the Impala Cookbook 🙂 Would love to see an updated presentation like this. Thanks Mostafa!

Your voice is sexy

*Promo SM*

I like how the word MOAR is the most ENLARGED TEXT lollll

we need MOAR talks like this! amazing!

😠 *promo sm*

Sounds cool

Very informative talk, I wonder if newer Transformer based models have been applied to sequential recommendations task (I've seen some papers but not public talks), would really appreciate if anybody shared ☺

Great presentation; is it possible to implement it in sequence log data clustering?

Can I get more information on this ...

Very clear explanation, I'm not a native english speaker but I understood everything clearly.

Awesome

nice explanation and great work, may be this is what me and my team were looking for thanks man!

Olá, gostaria de saber se a Criteo trabalha no Brasil, usando plataforma de marketing digital, eu estou fazendo parte de uma plataforma que usa nome da criteo, porém estou com dúvida se realmente é da Criteo. está plataforma libera 50 anúncios para nós divulgar na rede. e os ganhos são em dólares.

Great Video and better than great even explained! Thanks

The video sound is pretty good, beyond my imagination

at 15:50 the matrix profile looks like it is not plotted as is - based on the second subplot, the Y axis is between 0.5 and 2.5 whereas the matrix profile has values ranging from 1.4 to 14.1- what kind of processing has been done on the matrix profile before plotting it?

Super insightful!

16:22 .. I thought that guy was asleep lol

Can this be done using SPSS?

Your book is awesome everyone is talking about that book

Hope u earn 2000 pips a day

robust approach simplified.

Thanks God

Doneanotherartsdotsnotsdotsdots

13:45 Did the cameraman intentionally zoom in on that guy sleeping??

This is the best and easy way to find an anomalies but it's effective thanks Sean law introducing open source library stumpy

33:45 now that basically tells everything about this lecture

Thank you for explaining it well. I was searching for something with respect to biological data and may be I can use it. Great talk..

Internal Data University is for Spotify employees? If not, how do I apply?

wtf is this ? where to look at the slide? whole point of video is lost.... shit..

This guy knows what he is talking. However, his audience there and me be like, wtf is this.

Какой хороший, съела бы зацёмкала

11:50 The great engineer asks us to click star at github. It will make his day :) github.com/spotify/scio

Great talk! wonder where to have a look at the slides.