Top news of the week: 27.08.2020.

#explainability #explainableAI #observability #AI #Research #MachineLearning #NeurIPS2020 #tinyml #ECCV2020

Research

On Aug 23, 2020
@alexjc shared
RT @jeremyphoward: Today is https://t.co/GEOZuodrZj’s biggest day in our four year history. We are releasing our new course, fastai v2, fastcore, fastscript, fastgpu, and the official launch of the book. 1/ https://t.co/odaoVM75o0
Open

fast.ai releases new deep learning course, four libraries, and 600-page book

fast.ai releases new deep learning course, four libraries, and 600-page book

We are releasing: fastai v2: A complete rewrite of fastai which is faster, easier, and more flexible, implementing new approaches to deep learning framework design, as discussed …

On Aug 25, 2020
@peteskomoroch shared
RT @fiddlerlabs: Thrilled to announce a strategic investment from @amazon's @alexadevs. This collaboration will help to advance actionable insights in AI, furthering our goal of democratizing responsible AI development: https://t.co/Vnkwoeilj2 #explainableAI #observability #explainability
Open

Fiddler Secures Strategic Investment from Amazon Alexa Fund to Accelerate AI Explainability

Fiddler Secures Strategic Investment from Amazon Alexa Fund to Accelerate AI Explainability

Fiddler Secures Strategic Investment from Amazon Alexa Fund to Accelerate AI Explainability. The investment will help to enhance explainable AI and ML monitoring capabilities with the goal ...

On Aug 25, 2020
@clmt shared
RT @mirandaconrado: Along with our series A, I wrote a post on the MLOps challenges we have identified with customers. https://t.co/RHnG3Q1eOU Shoutout to my ex-colleagues @nkoumchatzky @clmt @tim_zaman @zakattacktaylor @yz @nimalan @pzdk @nojeshua and others who helped me mature the mental model
Open

The third wave of operationalization is here: MLOps

The third wave of operationalization is here: MLOps

Most companies fail at extracting value from using machine learning models in products because they can't operationalize models properly.

On Aug 25, 2020
@jaykreps shared
32/ September: Global https://t.co/qg4hmKXK0U
Open

Project Metamorphosis Month 5: Global Event Streaming in Confluent Cloud

Project Metamorphosis Month 5: Global Event Streaming in Confluent Cloud

We’re excited to announce a step-change release that makes event streams in Confluent Cloud 100% global. Link Kafka clusters anywhere in the world efficiently and with strong delivery ...

On Aug 25, 2020
@AlisonBLowndes shared
RT @Synced_Global: Sepp Hochreiter on Parallels Between Attention Mechanisms and Modern Hopfield Networks #MachineLearning #Research #AI https://t.co/xUmXUupX4C
Open

Sepp Hochreiter on Parallels Between Attention Mechanisms and Modern Hopfield Networks

Sepp Hochreiter on Parallels Between Attention Mechanisms and Modern Hopfield Networks

Researchers argue that “attention mechanism is the update rule of a modern Hopfield network with continuous states.”

On Aug 24, 2020
@kchonyc shared
RT @FeryalMP: @neuro_data @KordingLab @tdverstynen @danilobzdok @lakens @TonyZador @ylecun @criticalneuro @IrisVanRooij @neurograce @achristensen56 @AudeOliva @BlackInNeuro @BlackWomenInAI @tyrell_turing We'd love to see submissions addressing this question to our workshop on Biological and Artificial Reinforcement Learning at #NeurIPS2020. We'll share our CFP soon! :) https://t.co/yKWqCOQ3qD
Open

Workshop on biological AND artificial reinforcement learning

Workshop on biological AND artificial reinforcement learning

Reinforcement learning (RL) algorithms learn through rewards and a process of trial-and-error. However, artificial agents still struggle with a number of difficulties, such as …

On Aug 26, 2020
@petewarden shared
Some very cool work from Arm Research that combines Winograd transforms with quantized arithmetic for some big #tinyml performance improvements: https://t.co/DSKSIOxzXP
Open

Reducing the Cost of Neural Network Inference with Residue Number Systems

Reducing the Cost of Neural Network Inference with Residue Number Systems

The size and computational complexity of neural network models continues to grow exponentially. However, the increase in computational requirements presents a major challenge to their ...

On Aug 25, 2020
@MSFTResearch shared
Microsoft is a proud Diamond sponsor of #ECCV2020. Attendees can watch sessions & chat with researchers in our virtual booth: https://t.co/YW9xeJof4A Check out @mapo's tutorial on the new “Research Mode” capability of HoloLens 2 & SDK of Kinect for Azure: https://t.co/oDipKuzGOT
Open

Microsoft HoloLens 2 and Kinect for Azure as tools for computer vision research – Tutorial @ ECCV 2020

Microsoft HoloLens 2 and Kinect for Azure as tools for computer vision research – Tutorial @ ECCV 2020

Microsoft HoloLens 2 and Kinect for Azure as tools for computer vision research – Tutorial @ ECCV 2020