Day 55: A "Cookbook" for GAN research

New blog post discussing DeepMind and OpenAI's work on solving the first level of Montezuma's Revenge using Deep RL, and why that isn't necessarily as exciting as it seems. Would love to hear other's thoughts. https://t.co/LpXOIRZa7B

— Arthur Juliani (@awjuliani) July 13, 2018

New work from our group @GoogleAI - A practical “cookbook” for #GAN research in 2018: https://t.co/ZSIZQILWG3

Code & pre-trained TensorFlow Hub models: https://t.co/iwnSqznb3J

Attending #ICML2018 ? Visit our poster on Saturday (Reproducibility Workshop)

— Karol Kurach (@karol_kurach) July 13, 2018

The latest from the team that brought you "Are GANs created equal?" The new cookbook now endorses some GAN techniques as better than others. https://t.co/V6KDdqXaGM

— Ian Goodfellow (@goodfellow_ian) July 13, 2018

Training and evaluation code for "Adversarial logit pairing" paper is released: https://t.co/Kz8P3L8Cv7@goodfellow_ian @harinidkannan

— Alexey Kurakin (@alexey2004) July 13, 2018

`We are happy to announce our upcoming book "R Markdown: The Definitive Guide", authored by @xieyihui @fly_upside_down @StatGarrett and published by @CRCPress: https://t.co/ncKWmgKW5o pic.twitter.com/rpWE3iHJtz

— RStudio (@rstudio) July 13, 2018

There are some fantastic links on neural network acceleration hardware in this CS217 syllabus: https://t.co/f056QIt2AT - thanks @perdavan and Kunle Olukotun!

— Pete Warden (@petewarden) July 13, 2018

A very comprehensive paper for those into kernel methods and GPs.

Gives deep insights into the nature of our assumptions when modelling.

Thanks Motonobu Kanagawa, Philipp Hennig, @sejDino and Bharath K Sriperumbudur!https://t.co/UfNBoSHixu

— Neil Lawrence (@lawrennd) July 14, 2018

This week's #KernelAwards winner uses a pre-trained VGG16 model with attention to classify cardiomegaly from chest x-rays: https://t.co/Z4NwkMwoSa pic.twitter.com/7GBQk4nXB3

— Kaggle (@kaggle) July 13, 2018

A user just submitted an example visualization to Altair that I had no idea was even possible https://t.co/hZyh2WXHja pic.twitter.com/rIZj2S6EFS

— Jake VanderPlas (@jakevdp) July 13, 2018

#broom 0.5.0 is headed to CRAN on Monday (breaking changes!)

- Tidiers now return tibbles
- New vignettes and more thorough documentation
- Literally more bugfixes than I can count
- Tons of contributions from fantastic #rstats community membershttps://t.co/YBhjXeJy0T pic.twitter.com/C0rytRttUH

— alex hayes (@alexpghayes) July 13, 2018

ggspatial 1.0 is now on CRAN! @openstreetmap tiles, automatic spatial-aware scale bars and north arrows (with multiple styles thanks to @brentthorne18!), and fast (ish) RGB raster layers for #gis in #ggplot2 #rstats! https://t.co/XnTLa7dfm5 pic.twitter.com/0B1RSsvxmV

— Dewey Dunnington (@paleolimbot) July 13, 2018

https://t.co/cAsWniw70d is too cool.

I've updated my #DataScience & #MachineLearning projects repo so you can run my "Example #Python Machine Learning Notebook" directly in your browser, no installation required: https://t.co/amYzf1e66Q

Will update other projects as I get time. pic.twitter.com/5m18xpcc2K

— Randy Olson (@randal_olson) July 13, 2018

This is so cool. Automatically create a @Docker from your @github repo and create a @ProjectJupyter notebook too! Great idea to help with code lifespan! pic.twitter.com/sK8Z1GtJdm

— Olivia Guest (@o_guest) July 14, 2018