Day 50: Make Sad Plots Better

New piece about a direction I'm super excited about: NLP's ImageNet moment has arrived @gradientpub https://t.co/Kb3szva7tC

— Sebastian Ruder (@seb_ruder) July 9, 2018

Do check out today's other perspectives: on AI regulation by @MelodyGuan (https://t.co/ieF6NV6gS2) and on the flaws of RL by @andrey_kurenkov (https://t.co/ieGSIc477W). If you have sth to share, reach out to @gradientpub here: editor (at) thegradient (dot) pub

— Sebastian Ruder (@seb_ruder) July 9, 2018

1/4
10 days ago, when I decided to go on Twitter,
I did not know what I am getting into. Today,
with over 5,000 followers, I feel a sense of
obligation to give you a progress report, and to
reflect on this experience. This will probably take
4 Tweets - stay tuned #bookofwhy

— Judea Pearl (@yudapearl) July 8, 2018

2/4
First, I find the intellectual exercise of
compacting pages of technical analysis into
3-4 sentences challenging, educational and empowering.
Second, the genuine quest for understanding I
see among you gives me the hope that my primary goal
of writing the Book is realizable.

— Judea Pearl (@yudapearl) July 8, 2018

3/4
My goal was and is: the democratization of causal
inference . By that I mean empowering
students and researchers to understand causal inference
on their own, without waiting for professors or
journal editors to catch up.

Toward that end, I will continue to bring up ...

— Judea Pearl (@yudapearl) July 8, 2018

4/4
for discussion topics that have been ignored or suppressed by
mainstream literature and, simultaneously,
try to be as flexible, honest and open-minded
as I can in responding to your questions and comments.
Unortunately, I can only Tweet once or twice
a day. #bookofwhy

— Judea Pearl (@yudapearl) July 8, 2018

I've seen this image floating around, being used to claim that Fortnite is taking over Twitch - stealing viewers and streamers from other games. However, this doesn't seem to be true pic.twitter.com/vul3bVPuow

— Freya Holmér (@FreyaHolmer) July 7, 2018

Here we can see that the amount of channels streaming League, Overwatch and even its competitor PUBG, actually remain largely unaffected by the massive amount of Fortnite pic.twitter.com/V4pihfOaRf

— Freya Holmér (@FreyaHolmer) July 7, 2018

“The actual practice of AI research is more like architectural design than like electrical engineering.” https://t.co/1QQV8c3rmg

— hardmaru (@hardmaru) July 9, 2018

Architects build models from cardboard; AI researchers build them from code. These prototypes are not engineering models, subjected to serious real-world testing. They’re just “sketches” to give a sense of how something might work.

— hardmaru (@hardmaru) July 9, 2018

To develop expertise in AI, you can’t just read papers; you have to read other people’s code. And you can’t just read it, you have to re-implement it. Part of your understanding is gained only through the practice of coding itself.

— hardmaru (@hardmaru) July 9, 2018

Best paper award at #ICML2018 main idea: identify a class of adversarial defenses, called "Obfuscated gradients", & develop attack techniques to thwart this defense. Circumvents 6 of 9 defenses presented at #ICLR2018.
Paper: https://t.co/Z40rE2qKIx
Code: https://t.co/uDZxycCDN7

— Reza Zadeh (@Reza_Zadeh) July 8, 2018

Awesome!

A Reductions Approach to Fair Classification
https://t.co/0UvRIHQ3mS

- Simple mechanism to enforce expectation constraints
- Motivation: Fairness constraints (eg. demo parity, eq odds)
- Easy to impl: Add a loop around ANY existing classifier's train function! pic.twitter.com/aknO0xoyUs

— Tim Vieira (@xtimv) March 13, 2018

Code for this paper https://t.co/XolvOYZzXQ

- Code is very readable (Python) and includes some neat tricks
- Uses scikit-learn as the base algorithm https://t.co/fXOjPJSJDj

— Tim Vieira (@xtimv) July 8, 2018

#ipyleaflet 0.9.0 is out! With many new features by @Renou_Martin and @davidbrochart. Measure control, heatmap, velocity visualization.

Check out the maps visualization library for @ProjectJupyter https://t.co/4AHZ17bAyA pic.twitter.com/uO9iNga0fD

— Sylvain Corlay (@SylvainCorlay) July 8, 2018

just found this little gem https://t.co/4wqqbxsNRM in #edward from @dustinvtran and co....makes it super easy to work with different datasets

— Akash Srivastava (@variational_i) July 8, 2018

😻 I love this 📦 so much! Thx, @ewen_!
🎨 "ghibli: Studio Ghibli colour palettes" https://t.co/HsSDbmx3sO #rstats #dataviz pic.twitter.com/AAhtriEt3z

— Mara Averick (@dataandme) July 9, 2018

🎞 by the author, too!
"Take a Sad Plot & Make It Better" by @apreshill https://t.co/dGKk4hPIOl #rstats #dataviz #ggplot2 pic.twitter.com/olvsYrUhLe

— Mara Averick (@dataandme) July 8, 2018

It's extremely important that we anticipate & critically examine the harms of false positives in algorithmic decision-making. Not only in practice (like this real world scenario!) but in scholarship--if you classify things, let's see discussion of what happens when you're wrong. https://t.co/yhm0gBuVjz

— Casey Fiesler (@cfiesler) July 8, 2018

Why Sci-Hub is illegal, and what you can do about it https://t.co/ng4N791SzH pic.twitter.com/IcsfFavBfS

— Sci Hub (@Sci_Hub) July 7, 2018