Test set’s data distribution vs
— hardmaru (@hardmaru) October 12, 2018
Real world’s data distribution https://t.co/jYE1ULeids
Test set’s data distribution vs
— hardmaru (@hardmaru) October 12, 2018
Real world’s data distribution https://t.co/jYE1ULeids
Reminds me of the classic joke [setup: how do you become a millionaire doing X?] [punch: you start with two million]
— Zachary Lipton (@zacharylipton) October 11, 2018
Usually, "I have a social radar" means "I make categorical, discriminatory judgments about strangers after 30s of interaction, based on subjective personal preferences, including whether I like your face, and I am extremely confident about it"
— François Chollet (@fchollet) October 11, 2018
Exciting news in #GTC2018: @nvidia helping the open-source pydata ecosystem run on GPU, beyond deep learninghttps://t.co/iHLHv7asDu
— Gael Varoquaux (@GaelVaroquaux) October 10, 2018
Bummed to not be there, but today was also the first technical committee meeting for https://t.co/H57ZV1peaA (more on that later).
I knew organizing a useR! was a lot of work, but still wow 😲. Thanks @useR2018_conf for such a fantastic event! Lots of useful, concrete details here for people planning comparable events. https://t.co/9OIGszQFcw
— Jenny Bryan (@JennyBryan) October 10, 2018
Todai in AI https://t.co/q004XJtn3B
— fastml extra (@fastml_extra) October 9, 2018
Exercises to do when you feel nervous. Backstage at @TEDxSFbay #TEDxSanFrancisco pic.twitter.com/68BbS7LzKw
— Rachel Thomas (@math_rachel) October 9, 2018
Step 1 on the path to a nobel prize : learn python 3.6 f-strings https://t.co/FJO99BV8NZ
— Jeremy Howard (@jeremyphoward) October 9, 2018
Our @fastdotai international fellows are spreading their knowledge to nobel laureates now! :D https://t.co/cUPwu0vzML
— Jeremy Howard (@jeremyphoward) October 9, 2018
Hottest programming skills in 2018:
— Vicki Boykis (@vboykis) October 9, 2018
5. Fixing git merge conflicts
4. Correctly mapping ports in Docker containers to host machines
3. Getting info from AWS documentation
2. Pulling summary stats from a data stream
1. Turning any of the above into a conference talk about AI
The Economist is now on GitHub, and, looks like, uses R and Jupyter. https://t.co/GtCTRma3Af
— Vicki Boykis (@vboykis) October 9, 2018
Alarming, The things that biostatisticians are asked to do. https://t.co/plx7tbiJTFhttps://t.co/88IrIAyWBo@AnnalsofIM @ACPinternists cc @f2harrell @EdwardTufte pic.twitter.com/8PxaJNDcTE
— Eric Topol (@EricTopol) October 8, 2018