Mic.ai @mic_dot_ai

It’s an honour to be listed among these spectacular female powerhouses as one of 100 ‘brilliant women in AI ethics’. @MullerCatelijne @ruchowdh @jovialjoy @terahlyons @drfeifei @SophieCFischer @RosieCampbell @frossi_t etc etc 💪

Dr. Joy Buolamwini @jovialjoy

7 years ago

It’s an honour to be listed among these spectacular female powerhouses as one of 100 ‘brilliant women in AI ethics’. @MullerCatelijne @ruchowdh @jovialjoy @terahlyons @drfeifei @SophieCFischer @RosieCampbell @frossi_t etc etc 💪

1 18 33 0 3

Very excited to be giving a tutorial on Deep RL at @NASA next couple of days, and very curious about their use cases!

Scalable Deep Reinforcement Learning for Robotic Manipulation ai.googleblog.com/2018/06/scalab… hand-designed init -> 580k grasp attempts on 7 robot arms over 4 months (raw monocular RGB camera input) -> 1.2M param net -> 96% successful test set grasps

Google Cloud TPUs now offer preemptible pricing at ~70% off the reserved instance pricing. This means, for example, that you can train a ResNet-50 model for ~$7.50 instead of $25, or a Transformer neural translation model for ~$13 instead of $41. See: cloudplatform.googleblog.com/2018/06/Cloud-…

Graph Neural Networks / Relational Networks are models worth studying. We wrote a pretty comprehensive review about them which I hope you will find helpful (code forthcoming!). arxiv.org/abs/1806.01261

TensorFlow Serving has released version 1.8.0 & now supports a RESTful API out of the box! You can now easily make classification, regression & prediction requests to your TensorFlow models using JSON objects. Learn more here ↓ bit.ly/2y2xR5N

FloydHub has released an integrated developer environment for deep learning, based on Jupyter. Looks very nice! I like how user-focused FloydHub seems to be. blog.floydhub.com/workspaces/

Just finished reading through DeepMind's / @PeterWBattaglia et al.'s new review/position paper on graph neural nets: arxiv.org/abs/1806.01261 Timely and highly relevant contribution to the field IMO, couldn't agree more with their motivation for this class of models

This week on #AskTensorFlow, learn about suitable datasets for TensorFlow beginners, retraining models for object recognition, advice for those interested in getting started in AI/ML, and the use of Cloud TPUs. Watch here → bit.ly/2Jx3dG5

New release of Keras coming this week. 2.2.0. Lots of important updates. Excellent work from contributors -- open-source keeps amazing me.

Been working with Knowledge Graph Embeddings recently e.g. for Link Prediction (arxiv.org/abs/1703.08098) and with graph databases (neo4j). Also been reading about more general techniques on graphs e.g. thegradient.pub/structure-lear… arxiv.org/abs/1709.07604 Graphs are the future.

Denny Britz @dennybritz

7 years ago

Been working with Knowledge Graph Embeddings recently e.g. for Link Prediction (arxiv.org/abs/1703.08098) and with graph databases (neo4j). Also been reading about more general techniques on graphs e.g. thegradient.pub/structure-lear… arxiv.org/abs/1709.07604 Graphs are the future.

1 32 124 0 10

Really looking forward to @CheltSciFest this week. I’m doing events on quantum entanglement, wormholes, dark matter and artificial intelligence (with @demishassabis). Oh, and recording a Life Scientific with @closefrank. I think I have everything covered.

This conveyor belt can move in any directions V/@TechInsider @PatrickGunz_CH @rajat_shrimal @ipfconline1 @SpirosMargaris @ahier @HITpol @EvanKirstel @JacBurns_Comext @psb_dc @SpirosMargaris @guzmand @LouisSerge @chboursin @mallys_ @jerome_joffre @sebbourguignon @helene_wpli

Interesting interview with Andrew Ng 7wdata.be/deep-learning/…

Simone is out of surgery and her doctors are very pleased with how everything went. She woke up long enough to make an inappropriate joke... so things are looking good.

Google’s Rules of ML for products: developers.google.com/machine-learni… “Most of the problems you will face are, in fact, engineering problems. Even with all the resources of a great machine learning expert, most of the gains come from great features, not great machine learning algorithms”

Looking forward to presenting on "Vision-based Robot Control with Deep Learning" at the Vision-based Robotic Control tutorial tomorrow afternoon at #ICRA2018! Tutorial co-organized with Francois Chaumette, Robert Mahony, Fabien Spindler. Room M3. team.inria.fr/rainbow/icra18…

Fantastic endorsement of our recent grid cell paper by Edvard Moser (co-discoverer of grid cells with May-Britt Moser and co-winners of 2014 Nobel Prize w/ John O'Keefe) in @F1000Prime f1000.com/prime/73319806… deepmind.com/blog/grid-cell…

Could dopamine play a different role in learning than we previously thought? Our new @NatureNeuroscience paper investigates how the brain learns to learn using a meta-reinforcement learning framework from AI: deepmind.com/blog/prefronta…

As ML evolves from experimentation to serving production workloads, the need to manage end-to-end training & production increases This session introduces TensorFlow Extended, TensorFlow Hub & new innovations in TensorFlow Serving! Watch #io18 live ↓ goo.gl/sL3q3e