Still the best learning path for machine learning:

- A Mathematics Course for Political and Social Research
- Introduction to Statistical Learning
- Elements of Statistical Learning
- Deep Learning

Lots of directions you can take your learning after that.

— Chris Albon (@chrisalbon) January 29, 2021

South Korea is one of the few developed countries still reporting rabies cases (albeit extremely low numbers—no human infection since 2004 and no animal infection since 2014). Officials blame unvaccinated animals from North Korea crossing the porous DMZ. https://t.co/E7fkHPN0n0 pic.twitter.com/1RapcjqjZF

— Simon Kuestenmacher (@simongerman600) January 29, 2021

Three rules of thumb for machine learning:

1) be skeptical of good results
2) be skeptical of bad results
3) be skeptical of mediocre results

— Thomas (@evolvingstuff) January 28, 2021

Piece of advice that helped me: Pay attention to how you feel when interacting with folks. Do they make you feel excited? Smart? Stupid? Lazy?

One of the highest value things you can do for yourself is shadowbanning from your life folks who makes you feel worse about yourself.

— Chris Albon (@chrisalbon) January 28, 2021

There has been much interest in ML methods that generate source code (e.g. Python) from English commands. But does this actually help software developers? We asked 31 developers to use a code generation plugin, and found some interesting results: https://t.co/ifiG3EYK3J 1/7 pic.twitter.com/u0mgY0LSSj

— Graham Neubig (@gneubig) January 28, 2021

"The results ... suggest that decision-makers can actually rid themselves of guilt more easily by delegating to machines than by delegating to other people." 😬

"Hiding Behind Machines: When Blame Is Shifted to Artificial Agents," Feier et al.: https://t.co/fpNTNdkY7A

— Miles Brundage (@Miles_Brundage) January 28, 2021

High-Low Frequency Detectors — A new Distill article by @ludwigschubert, @csvoss, and @ch402. This is the fifth article in the circuits thread.https://t.co/b8Bfl0heOo

— Distill (@distillpub) January 27, 2021

Want Transformers to perform long chains of reasoning and to remember stuff?
Use Feedback Transformers.
Brought to you by a team from FAIR-Paris.@tesatory https://t.co/dpCzt2yiPf

— Yann LeCun (@ylecun) January 27, 2021

what a simple yet effective idea! :)

looking at it from the architectural depth perspective (https://t.co/DM4MvzqFQW by zheng et al.,) the depth (# of layers between a particular input at time t' and output at time t) is now (t-t') x L rather than (t-t') + L. https://t.co/GkkuMH7a5u pic.twitter.com/eXWeqoEKxx

— Kyunghyun Cho (@kchonyc) January 27, 2021

Have you heard about this nbdev thing but wondering what its really like to write code in it?

Watch me as I live-code a nbdev project from scratch 😆 👇✨

Thanks to @NorthwesternU for hosting the eventhttps://t.co/sOXb2unMGg

— Hamel Husain (@HamelHusain) January 27, 2021

Recent updates to the live transcription feature in the Google Translate app reduce revisions, providing a noticeable improvement to the user experience. Learn how masking and biasing enable high accuracy with low erasure and minimal lag at: https://t.co/9aV4pjcZj1 pic.twitter.com/ZsNObLw141

— Google AI (@GoogleAI) January 26, 2021

So cool to see Set Transformer being used for smart model ensembling. Kaggle competitions are becoming more interesting and Transformer models are getting more attention! https://t.co/6OKOutqOKQ pic.twitter.com/cMp2NSBCQu

— Alexandr Kalinin (@alxndrkalinin) January 26, 2021