Data science using data to answer questions. Your answers are only as good as your data and your questions.
— Brandon Rohrer (@_brohrer_) November 4, 2019
Data science using data to answer questions. Your answers are only as good as your data and your questions.
— Brandon Rohrer (@_brohrer_) November 4, 2019
One cool thing about writing a book on data science is that you automatically create your own corpus for analysis! The data is there for the taking! pic.twitter.com/1KWsQqvhiL
— Jacqueline Nolis (@skyetetra) November 4, 2019
The Compas Recidivism Algorithm:
— Rachel Thomas (@math_rachel) November 3, 2019
- it's no more accurate than random people (Amazon Mechanical Turk)
- it's a black box with 137 inputs but no more accurate than linear classifier on 2 vars
- Wisconsin Supreme Court upheld its use (it is still used in other states as well) pic.twitter.com/cMdGOkQPP5
Much of the excessive surveillance and control we fear and recoil from in authoritarian countries is wielded by companies in capitalist ones. They just wield it with less efficiency, more subtlety, and through legal finesse. Capitalism doesn't have morality baked in by default. https://t.co/wzKIoQ1vr8
— Smerity (@Smerity) November 3, 2019
One of the biggest sources of institutional decay is the lack of risk / benefit analysis. It's so easy to overcorrect after a catastrophe, and incur an invisible cost forever after. Do that for a couple of years, and before you know it you've outlawed innovation and productivity
— Florent Crivello 🌐 (@Altimor) November 2, 2019
I love this deeply and with all my heart. https://t.co/KIyBk5mhOQ
— Dieter Bohn (@backlon) November 1, 2019
Here's a look at ads for financial services like loans and banks targeted only to men and younger consumers: https://t.co/PclEjZ6dZh@jeremybmerrill & @johndetrixhe used the Facebook Ad Collector to see them. Facebook STILL doesn't publish targeting info: https://t.co/ZHEqU5qqs1 https://t.co/hEs8exkSQR
— Ariana Tobin (@Ariana_Tobin) November 1, 2019
OMG Just found this amazing extension of my sklearn cheat-sheet by @chris_bour:
— Andreas Mueller (@amuellerml) November 1, 2019
https://t.co/T5vZ1I5g13 pic.twitter.com/bbwRfiixif
In my experience, for selfish reasons alone, releasing a draft as soon as the paper is ready is absolutely the optimal strategy. Much of my best known work was public for *years* before it was official. Not-yet-accepted work was a big part of my job talk and even my tenure case.
— Arvind Narayanan (@random_walker) November 1, 2019
the reason I got into data science was because I (was an "analyst" who) wanted to move beyond just doing BI work; and the reason I got out of data science was because I was doing a lot of software development and discovered that I really really enjoyed it https://t.co/DUTKGbfqLD
— Joel Grus ♥️ 📓 (@joelgrus) November 1, 2019
3) Data Scientists who are stuck building data warehouses when their company really needed to hire data engineers first.
— Data Science Renee (@BecomingDataSci) November 1, 2019
In my limited experience every single data sci type has one of the following two gripes:
— JD Long (@CMastication) November 1, 2019
1) I'm stuck doing BI work and it's just glorified pivot tables and dashboards
- or -
2) I'm doing too much software development and I want to actually work harder on the data analysis. https://t.co/H9LWDl01Ny