And it would be a crime to leave #ICML2018 out.

It is held in Stockholm now, papers are available at: http://proceedings.mlr.press/v80/

#conference

Toward theoretical understanding of deep learning

http://unsupervised.cs.princeton.edu/deeplearningtutorial.html

#ICML2018 #tutorial #dl

Facebook Research at ICML 2018

Seventeen papers by #FAIR at the #ICML2018

https://research.fb.com/facebook-research-at-icml-2018/

#Nvidia developed approach for cleaning noisy photoes. This comes extremely handy in the radiology.

Link: https://news.developer.nvidia.com/ai-can-now-fix-your-grainy-photos-by-only-looking-at-grainy-photos/?ncid=--45511

#dl #cv

ICML 2018 Accepted Papers Stats

Medium post with the meta statistics about accepted papers.

https://medium.com/machine-learning-in-practice/icml-2018-accepted-papers-stats-339f1cad5a55

#ICML2018 #meta #statistics

Visitor's highlights on #ICML2018

[ICML 2018] Day 2 - Representation Learning, Networks and Relational Learning

https://gmarti.gitlab.io/ml/2018/07/11/icml18-day-2.html

Ian Goodfellow (Google Brain) will have a free webinar on Adversarial Machine Learning

https://event.on24.com/eventRegistration/EventLobbyServlet?target=reg20.jsp&partnerref=twitterShareFromReg&ms=1531251848082&eventid=1633807&sessionid=1&key=8B7A8F4B65B54C35752F8A6FE23F641A&regTag=&sourcepage=register

#webinar #google #gbrain

Very thoroughful notes on #ICML2018

https://david-abel.github.io/blog/posts/misc/icml_2018.pdf

Learning World Models: The Next Step Towards AI

Video of Yann LeCun's keynote speech. First-hand expectations of closeness of real AI. Spoiler: still not even near yet.

https://www.facebook.com/722677142/posts/10155402036352143/

Takeaways from Netflix’s Personalization Workshop 2018

https://medium.com/rtl-tech/my-takeaways-from-netflixs-personalization-workshop-2018-f564a19437b6

#recommendation #netflix #summary

A cool paper from Facebook AI (not from FAIR!) about detecting and reading text in images, at scale. 

This is very useful for detecting inappropriate content on Facebook.

The system uses R-CNN/Detectron for detecting lines of text.

The OCR uses a ConvNet applied at the level of a whole line trained with CTC.

This concept of applying a ConvNet on a whole line of text, without prior segmentation, has roots in the early days of ConvNets, for example with this NIPS 1992 paper:
"Multi-Digit Recognition Using a Space Displacement Neural Network"
by Ofer Matan, Chris Burges, Yann LeCun and John Denker.

Link: https://papers.nips.cc/paper/557-multi-digit-recognition-using-a-space-displacement-neural-network
Youtuve video with short explanation: https://youtu.be/yl3P2tYewVg

#ocr #cv #dl #rnn #facebook #yannlecun #video

Cancer metastasis detection with neural conditional random field (NCRF)

Github: https://github.com/baidu-research/NCRF?utm_source=telegram&utm_medium=opendatascience

#Baidu #Cancer #Segmentation #cv #DL

XGBoost deployment made easy

Rather obscure way to serve XGBoost models, but still works.

Link: https://towardsdatascience.com/xgboost-deployment-made-easy-6e11f4b3f817

#production #xgboost #deployment

Series of articles «Machine Learning Kaggle Competition» in three parts:

1. Part I: https://towardsdatascience.com/machine-learning-kaggle-competition-part-one-getting-started-32fb9ff47426
2. Part II: https://towardsdatascience.com/machine-learning-kaggle-competition-part-two-improving-e5b4d61ab4b8
3. Part III: https://towardsdatascience.com/machine-learning-kaggle-competition-part-three-optimization-db04ea415507

#kaggle #guide

friend: yo dude, wanna drink?

me: nah man, that stuff kills brain cells.

friend: you say its killing brain cells but i say its just real life dropout to prevent overfitting

OpenAI bot defeated 5 human top 99.95 percentile DotA 2 players.

https://twitch.tv/openai

#dota #rl #openai