Kaggle's much anticipated "State of Machine Learning and Data Science" survey results are out! 🙌
https://t.co/4fRhYPzES5
I have been closely following this major survey for over half a decade now. A few things that stood out for me in 2021 relative to the last 2 years. 👇🧵
@alvaroviebrantz Massa. Também estou querendo fazer isso, adicionar uns HDs velhos, e instalar https://t.co/e3SHPcsUL9 ou algo do tipo pra configurar storage classes.
I wonder how long it will take for ARM-based systems to become commonplace and be fully supported by cloud computing providers and data management platforms.
It is pretty amazing what people are doing at home with small, inexpensive single-board computers (all running on ARM-based CPUs, the same architecture used in most smartphones):
https://t.co/4LqAw0BgaX
You wire together a few Raspberry Pis computers with some hard-drives, install an open-source distributed block-storage system (https://t.co/e3SHPcsUL9), and you get a scalable cluster with a 55TB of replicated storage accessible with an S3-like API or distributed file system.
Besides clearly not doing enough to help to combat misinformation and increase political ads transparency, Facebook now wants to stop independent NYU researchers from doing so.
This is a disturbing development.
Last Friday, Facebook sent a cease-and-desist letter to two NYU researchers (@LauraEdelson2 and Damon McCoy), demanding that they shut down their research into political ads and disinformation on FB’s platform.
https://t.co/ksBVmqV27K
The Bias-Variance Trade-Off & "DOUBLE DESCENT" 🧵
Remember the bias-variance trade-off? It says that models perform well for an "intermediate level of flexibility". You've seen the picture of the U-shape test error curve.
We try to hit the "sweet spot" of flexibility.
1/🧵
@ananoterminal @mariomarroquim Mário, fiquei curioso: tem algum motivo específico pra ter escolhido auto-sklearn ao invés de outros frameworks de AutoML? Performance, facilidade, compatibilidade com scikit-learn, ou só por acaso mesmo?
@ananoterminal Se tiver interesse em conversar sobre alguns projetos de pesquisa na NYU e academia em geral, me chama no DM. Talvez eu possa te ajudar.