State-of-the-art (SOTA) collection of Paperswithcode

A great site, worth spreading word about: Papers With Code now includes 950+ ML tasks, 500+ evaluation tables (including SOTA results) and 8500+ papers with code. Probably the largest collection of NLP tasks, including 140+ tasks and 100 datasets.

Link: https://paperswithcode.com
Sota link: https://paperswithcode.com/sota

#Meta #collection #sota #useful

Manifold: A Model-Agnostic Visual Debugging Tool for Machine Learning at Uber

Seesm like there is no week without any news from #Uber engineering team. This time Uber built Manifold, a model-agnostic visualization tool for #ML performance diagnosis and model debugging, to facilitate a more informed and actionable model iteration process.

Link: https://ubere.ng/2Hac0O8

#Pipeline #administration

​​Implementation of character based convolutional neural network

A #PyTorch implementation of Character Based ConvNets for text classification published by Yan LeCun in 2015 is open-sourced on. Many training features and hacks are implemented.

Link: https://github.com/ahmedbesbes/character-based-cnn

EE-559 – DEEP LEARNING (SPRING 2019)

Deep learning course covering the main deep learning tools and theoretical results, with examples in the #PyTorch framework.

Taught by François Fleuret from École Polytechnique Fédérale de Lausanne, Switzerland.

Link: https://fleuret.org/ee559/

#DL #course #learnhardgopro

​​Neural Networks seem to follow a puzzlingly simple strategy to classify images

Interesting article on how actually #NN see images and what helps to distinct different classes.

Link: https://medium.com/bethgelab/neural-networks-seem-to-follow-a-puzzlingly-simple-strategy-to-classify-images-f4229317261f

#BagNet #ResNet #Dl #CV

What do you get if you apply Data Science to Boxing ?

Been waiting for a long time to post something like this. Meet: BoxingBot, a robot, which uses #CV and various motion detectors to evade punches. It also can track statistics of hits. This is how #DL and #ML can improve any area of our lives, including such classy and old school, like boxing.

​​Machine Learning-Powered Search Ranking of Airbnb Experiences.

Post on how #AirBnB DS team built custom search, including notes on how they approached problem and what business results they achived.

Link: https://medium.com/airbnb-engineering/machine-learning-powered-search-ranking-of-airbnb-experiences-110b4b1a0789

#ranking #search #reallifeds #production

Text-based game to feel yourself in not so distant future where real AI exists.

We all now, how DS and #AI products affect our life and how they actually influence our lifestyle. But in the future our relationship will become more tense. It is rather nice experience to play this game, not only because it reminds of old-school games, but also because it looks shookingly possible.

Link: https://www.theverge.com/2019/1/31/18140796/wake-word-algorithm-text-game-ai-artificial-intelligence

#interactive #Meta

A Lyapunov-based Approach to Safe Reinforcement Learning

Research that develops RL algorithms based on the concept of Lyapunov functions. This work represents a step toward applying RL to real-world problems.

ArXiV: https://arxiv.org/abs/1805.07708

#RL #Lyapunov #facebook

​​Why Financial Planning is Exciting… At Least for a Data Scientist

Great introduction into the finance world and what data scientist can lack diving into the topic.

Link: https://eng.uber.com/financial-planning-for-data-scientist/

#Financial #statistics #Uber

​​IBM aiming for 1000x improvement in AI computation over the next 10 years.

How DS / AI sphere will develop according to #IBM prediction.

Link: https://www.nextbigfuture.com/2019/02/ibm-investing-2-billion-in-an-ai-center-and-targets-1000-times-improvement-by-2029.html

​​TOWARDS FEDERATED LEARNING AT SCALE: SYSTEM DESIGN

Google published how they do #FederatedLearning at scale on tens of millions of mobile phones. This is about training model on decentralized data.

ArXiV: https://arxiv.org/pdf/1902.01046.pdf

#Google #Privacy

​​A Style-Based Generator Architecture for Generative Adversarial Networks

Code and pre-trained models for Style-GAN paper.

Github: https://github.com/NVlabs/stylegan
ArXiV: http://stylegan.xyz/paper

#GAN #CV #DL

​​Intuitive Visualization of Outlier Detection Methods

Mostly with Python Outlier Detection.

Link: https://www.kdnuggets.com/2019/02/outlier-detection-methods-cheat-sheet.html

#outliers #dataviz #visualization

​​Advanced Technologies for Detecting and Preventing Fraud at Uber

Uber’s article on how they detect and prevent fraud, analyzing GPS traces and usage patterns to identify suspicious behavior.

Link: https://eng.uber.com/advanced-technologies-detecting-preventing-fraud-uber/

#geodata #Uber #fraud #GPS

It’s better to be late, than to miss amazing opportunite to take part in not-official, but supervised online course of Deep Learning by CMU.

Course has already started, but you still can hop on the departing train to learn (or review) deep learning foundations and have some practice.

Content, quizes, practice exercises are in English.

This initiative was brought by editor of russian-speaking @powerofdata channel.

Link: https://dlcourse.ru
Original course page (with mention of aforementioned initiative): http://deeplearning.cs.cmu.edu

A new ELF OpenGo bot and analysis of historical Go games


Facebook AI Research shared new features & research results related to ELF OpenGo, including an updated model that was retrained from scratch. Bonus: Windows executable version of the bot, and a unique archive analyzing 87K professional Go games.

Link: https://ai.facebook.com/blog/open-sourcing-new-elf-opengo-bot-and-go-research/

#facebook #gogame

​​#Infographic : 8 ways AI is Transforming #healthcare

Application of #StyleGAN: site, which generates new people every page refresh.

Link: https://thispersondoesnotexist.com