Day 54: ICML2018 and UseR2018

my new post: What do machine learning practitioners actually do? https://t.co/fNOKVJGISr

— Rachel Thomas (@math_rachel) July 12, 2018

"Is cleaning data really part of ML? Yes. [..] The process of cleaning a data set and training a model are usually interwoven: I frequently find issues in the model training that cause me to go back and change the pre-processing for the input data." https://t.co/GenYS5Sif6

— Monica Rogati (@mrogati) July 12, 2018

Everyone gets burned out, even @gvanrossum. https://t.co/bwwJuNd0Bk

— Kenneth Reitz (@kennethreitz) July 12, 2018

whoa. 😳 pic.twitter.com/iVteqltNtn

— 👩‍💻 DynamicWebPaige @ SciPy2018 🔬🐍 (@DynamicWebPaige) July 12, 2018

As Python becomes the most popular programming language out there, its creator Guido van Rossum is taking a "permanent vacation" from his role of Benevolent Dictator for Life https://t.co/RYQ5Ld0oj5

— Parker Higgins (@xor) July 12, 2018

umap: Uniform Manifold Approximation and Projection for dimensionality reduction. #Python #DataScience #MachineLearning

Claims to have several advantages over tSNE.https://t.co/0N06f8ktcn pic.twitter.com/Bh5ogX6dTQ

— Randy Olson (@randal_olson) July 13, 2018

UMAP version 0.3 is now available. You can now add new data to an existing embedding, embed using labelled data, or use both features for metric learning. Documentation is on readthedocs: https://t.co/ZFaOHrPti4. pic.twitter.com/CqiuZFbmCQ

— Leland McInnes (@leland_mcinnes) July 12, 2018

Slides for my @SciPyConf talk this morning on UMAP can be found here: https://t.co/XsZVuRxb5a

— Leland McInnes (@leland_mcinnes) July 12, 2018

Universal Transformers propose to augment Transformers with Recurrence in depth and Adaptive Computation Time. This model outperforms Vanilla Transformers in MT / bAbI / LA / LTE.

Paper: https://t.co/U2YAeuO6EO
Code: Soon in https://t.co/KSuQAkn5Jh pic.twitter.com/lCKfsEAswG

— Oriol Vinyals (@OriolVinyalsML) July 12, 2018

Back in Sydney after a great trip to Brisbane. Just wanted to say thank you to everyone at #useR2018, and especially to @visnut for inviting me! The slides for my talk are here:https://t.co/Q6BMC5ZLXT

— Danielle Navarro (@djnavarro) July 13, 2018

Roger Peng’s #useR2018 keynote this morning resonates with me, as another long time user/developer/instructor. Useful, opinionated take on where we are now in #rstats and how we got here. @rdpeng https://t.co/bOLSoaFupd pic.twitter.com/ejc9yFYGVA

— Jenny Bryan (@JennyBryan) July 13, 2018

Here is the video for my keynote from #useR2018 on teaching R to new users. https://t.co/KUrG097D7D

— Roger D. Peng (@rdpeng) July 15, 2018

🖤ed @rdpeng's #useR2018 keynote? Pairs nicely with 👇
🎬 "John Chambers' Keynote Speech from useR! 2014"https://t.co/TIaSFoGBRj via @wwwDSLA #rstats pic.twitter.com/ytV4pvERP7

— Mara Averick (@dataandme) July 13, 2018

📽 #useR2018 sneak-peak:
⚡️ "fasster: Forecasting multiple seasonality w/ state switching" by @mitchoharawildhttps://t.co/TL0rc18f0v #rstats pic.twitter.com/TNdMuwzeVy

— Mara Averick (@dataandme) July 12, 2018

Holy. Shit. Expressive, interesting, and lengthy piano music generations. I genuinely got goosebumps listening to these. As far as I'm concerned, this is a huge step forward for creative ML. Cheers Anna Huang. Paper: https://t.co/D50bdGHhxO #icml2018 pic.twitter.com/WjECu2h3Pl

— James Owers (@jamesowers) July 14, 2018

Overcoming Catastrophic Forgetting with Hard Attention to the Task #ICML2018
By @serrjoa @Surisdi @nkundiushuti @alexk_z

Reducing cutting forgetting rates by 45 to 80%.
As a by-product hard attention masks enable weights pruning and compression.https://t.co/XGMTlYAGoj pic.twitter.com/4ISbGpzShi

— ML Review (@ml_review) July 15, 2018

Hear some AI generated music at our talk and poster today at #ICML2018! Talk is at 2:50 in Victoria. Poster is #175 and 👇@jesseengel @colinraffel @fjord41 @douglas_eck pic.twitter.com/ufSesFVSTV

— Adam Roberts (@ada_rob) July 12, 2018

Check out the Image Transformer by @nikiparmar09 @ashVaswani others and me. Talk at 3:20p @ Victoria (Deep Learning). Visit our poster at 6:15-9:00p @ Hall B #217! https://t.co/stuwPR1fGt pic.twitter.com/i86usHSZIV

— Dustin Tran (@dustinvtran) July 12, 2018

Love this paper, it restores your faith in humanity.

Great challenge, great science, and brings together a number of ideas across different fields.

More of this type of work please! @dennisprangle #ICML2008 https://t.co/p4D6nx6TW0

— Neil Lawrence (@lawrennd) July 12, 2018

Slides of my talk at #ICML2018 on reproducible #MachineLearning workshop https://t.co/8PCk5cZlnJ https://t.co/cH3PhRZz4K #RML2018 @scikit_learn #opensource #openscience pic.twitter.com/CU22oMgsoS

— Alexandre Gramfort (@agramfort) July 14, 2018

How do we specify priors for Bayesian neural networks? Check out our work on Noise Contrastive Priors at the ICML Deep Generative Models workshop 11:40am+. @danijarh, @alexirpan, Timothy Lillicrap, James Davidson https://t.co/OS3gmKin9g pic.twitter.com/cNhfeCYuNb

— Dustin Tran (@dustinvtran) July 15, 2018

Plant seedling classification: a competition-winning approach using data augmentation and Keras https://t.co/uCFHmriodt

— Ben Hamner (@benhamner) July 13, 2018

Must-watch television: Bayesian Data Science Two Ways: Simulation and Probabilistic Programming, featuring @hugobowne and @ericmjl https://t.co/a6fVcLOOaz

— Chris Fonnesbeck (@fonnesbeck) July 12, 2018

For anyone interested, I'm (mostly done) writing a book which teaches Deep Learning using intuitive examples more than math. All code examples are on Github, written from scratch in Numpy

If you think this is your learning style, you can download it here: https://t.co/6KdGghoqLC

— Trask (@iamtrask) July 12, 2018

All of the available winners' solution from our ML competitions in one meta-kernel! Thank you @sudalairajkumar https://t.co/w5Y47Xd7sA

— Kaggle (@kaggle) July 12, 2018

Today we’re launching Seedbank, a place to discover interactive ML examples which you can run from your browser, no set-up required. Each example can be edited, extended, and adapted into your own project.

Read @mtyka's post for more info ↓ https://t.co/k1McWSm8PG

— TensorFlow (@TensorFlow) July 12, 2018

Did you know there's a package to do fuzzy joins in R? See fuzzyjoin by @drob https://t.co/beBwrpqMNN#rstats h/t @jakekaupp pic.twitter.com/TVL3nHXHsf

— Sharon Machlis (@sharon000) July 12, 2018

#Gensim 3.5.0 is out! 📚 MASSIVE documentation overhaul, consistent docstrings, new usage examples… (plus a bunch of new features, fixes and improvements). Enjoy~https://t.co/Lmch7ryVqw pic.twitter.com/ZQ14n4wev0

— Gensim (@gensim_py) July 6, 2018

Our group would like to make error messages easier for a user to act on. It’s hard, but developing a guide has been a productive first step. https://t.co/oTHQ2A8FLs https://t.co/tuZRMSI3a3

— Jenny Bryan (@JennyBryan) July 12, 2018