The top 12 top drugs for 2018 = ~$100,000,000,000
— Eric Topol (@EricTopol) October 20, 2019
$51.4 Billion for 6 #cancer drugs
$34.5 Billion for 3 immunology drugs
$12.9 Billion for 1 heart, 1 eye drug
(in the old days a blockbuster was $1B) pic.twitter.com/lfbOFSuFE3
It is not simply that medical studies do not "replicate well" They often contradict and reverse each other John's article in PLOS ONE https://t.co/gzikWLy6n8 has been viewed/downloaded an incredible 2.9 mill. times and been cited 3.3K times Can you argue with success? @MDiPaolaMD
— Spyros Makridakis (@spyrosmakrid) October 20, 2019
Fantastic piece by @math_rachel on the misuse of metrics in organizations, particularly ones that measure success by number of clicks and time spent on platform. https://t.co/H4gDs2scD1
— Vicki Boykis (@vboykis) October 19, 2019
Learners should be required to “shoulder surf” with experienced analysts to learn BOTH how things are done but also to realize pros also feel “wtf is going on?” a LOT when analyzing data. And we accidentally reinvent crap, and go down dead ends. Here’s a great ride along w Sean: https://t.co/TeMxR4BzlL
— JD Long (@CMastication) October 19, 2019
It's sadly not a CSV file, it's a "dat.gz" file and you need some XML schema file to read it in. In R you need the `ipumsr` package to actually load it (ughh). But about 10 mins later, I have a data frame! Game on. pic.twitter.com/Ip15Nxf6xV
— Sean J. Taylor (@seanjtaylor) October 19, 2019
I think this is an interesting topic but found this visualization hard to follow (no surprise if you've been reading my complaints about animated plots).
— Sean J. Taylor (@seanjtaylor) October 19, 2019
I have nothing to do tonight so i'm going to try to re-visualize this data. Starting a THREAD I'll keep updated as I go. https://t.co/aHl4S3L9iP
These are *actual* notes from Peter Thiel's class at Stanford on how to run startups, featuring VPs from Palantir and PayPal. And it gets worse from here. https://t.co/JJVSgDhjb5 pic.twitter.com/2cyh4YmfSb
— Kate Crawford (@katecrawford) October 19, 2019
The Data Scientist who rules the ‘Data Science for Good’ competitions on Kaggle 👏 | by @pandeyparul https://t.co/SsG0WEgvia
— Kaggle (@kaggle) October 18, 2019
Most popular posts on my blog! Have you read them yet?
— B Nalini (@purpultat) October 18, 2019
Deep Learning with https://t.co/473FAN6NMo: https://t.co/Hb7Jhuq4s4
Hitchhiker’s Guide to AI Ethics 3-part series:
1: https://t.co/mTePb0yNG5
2: https://t.co/XlY62X6R1v
3: https://t.co/7e869PFK1c#AIEthics #DeepLearning
💥 @LysandreJik just published a great comparison of inference speed & memory requirements for a wide range of recent NLP models:
— Thomas Wolf (@Thom_Wolf) October 18, 2019
- RoBERTa, Bert, GPT2, XLNet, DistilBert, CTRL...
- comparing PyTorch & TensorFlow 2.0
- on typical CPU and GPU setups
- summarizing good-practices👇 https://t.co/2G3cYaUozL
50 Things That Made the
— Tim Harford (@TimHarford) October 18, 2019
Modern Economy: "utterly addictive...a mesmerising mix
of history, marketing and psychology." - The Times.
Very perceptive review, I think. You can listen to episodes here: https://t.co/Bx3hOqZiuG
“Interpretability has become the chronic fatigue syndrome of AI” - @zacharylipton @ipam_ucla
— Kyle Cranmer (@KyleCranmer) October 17, 2019
It’s also a “Waste-basket taxon” like the antelope https://t.co/UCkZwTYJvO pic.twitter.com/J1yUT4YSeb
