Presentation streaming live from https://Neuralink.com at 8pm Pacific

Calculus on Computational Graphs: Backpropagation
Blog by Christopher Olah: https://colah.github.io/posts/2015-08-Backprop/
#Calculus #ComputationalGraphs #Backpropagation

A Beginner’s Guide to Python for Data Science
https://towardsdatascience.com/a-beginners-guide-to-python-for-data-science-60ef022b7b67

Generative models are an extension of Richard Feynman's quote, "What I cannot create, I don't understand."

📃 Understanding deep learning requires rethinking generalization (paper ref from beginning of the talk):

https://openreview.net/pdf?id=Sy8gdB9xx

Efficient Video Generation on Complex Datasets
pdf: arxiv.org/pdf/1907.06571
abs: arxiv.org/abs/1907.06571

This year I wrote a book teaching Deep Learning - it's goal is to be the easiest intro possible

In the book, each lesson builds a neural component *from scratch* in #NumPy

Each *from scratch* toy code example is in the Github below. https://github.com/iamtrask/Grokking-Deep-Learning

The Bach Doodle: Approachable music composition with machine learning at scale. arxiv.org/abs/1907.06637

Cross-Lingual Transfer Learning for Question Answering. arxiv.org/abs/1907.06042

Minimal Sample Subspace Learning: Theory and Algorithms. arxiv.org/abs/1907.06032

Hands On Bayesian Statistics with Python, PyMC3 & ArviZ
Gaussian Inference, Posterior Predictive Checks, Group Comparison, Hierarchical Linear Regression

https://towardsdatascience.com/hands-on-bayesian-statistics-with-python-pymc3-arviz-499db9a59501

MaskPlus: Improving Mask Generation for Instance Segmentation. arxiv.org/abs/1907.06713

Quant GANs: Deep Generation of Financial Time Series. arxiv.org/abs/1907.06673

Motion Planning Networks: Bridging the Gap Between Learning-based and Classical Motion Planners

Qureshi et al.: https://arxiv.org/abs/1907.06013

#Robotics #ArtificialIntelligence #MachineLearning

“What is Applied Category Theory?”

A collection of introductory, expository notes inspired. By Tai-Danae Bradley : https://arxiv.org/pdf/1809.05923.pdf) arxiv.org/pdf/1809.05923

#AppliedCategoryTheory #ACT #CategoryTheory

Awesome 👏🏻 news from Uber for releasing its code for conversational #AI
It’s called Plato Research Dialog System, and it was released in open source today on GitHub.
Github: https://github.com/uber-research/plato-research-dialogue-system
Blog: https://eng.uber.com/plato-research-dialogue-system/
Plato is designed for both users with a limited background in conversational AI and seasoned researchers. It has a clean and understandable design, integrating with existing deep learning and Bayesian optimization frameworks (for tuning the models), and reducing the need to write code.

(Microsoft) Obj-GAN Turns Words into Complex Scenes

TD;LR
The model is capable of generating relatively complex scenes based on a short phrase. The generator identifies descriptive words and object-level information to gradually refine the synthesized image.

Blog: https://medium.com/syncedreview/microsoft-obj-gan-turns-words-into-complex-scenes-5c6024f0f91d
Paper: Object-driven Text-to-Image Synthesis via Adversarial Training (CVPR 2019) : https://arxiv.org/pdf/1902.10740.pdf
code: https://github.com/jamesli1618/Obj-GAN

Preprocessing for deep learning: from covariance matrix to image whitening
https://hadrienj.github.io/posts/Preprocessing-for-deep-learning/

A very good ebook for ML
https://www.allitebooks.in/machine-learning-with-tensorflow/