Machine Learning (ML) Links

Very General
Machine learning for everyone (excellent; the first 2 videos are essential, but don't miss the spooky 'Neural net is trying to speak' near the end)
TensorFlow Playground ..... highly recommended
Timeline of machine learning
What is machine learning?
----- Carnegie Mellon University
----- IBM
----- MathWorks
----- MIT (David Sontag)
----- Oracle
----- Royal Society
----- Stanford (Andrew Ng)

General
Artificial intelligence weaknesses

AutoML code generation
Commercial software
----- Best machine learning software (reviews and comparison by g2.com)
----- Geophysical_Insights/Paradise (correlation-based ML for seismic interpretation)
----- Tibco/Spotfire (used by some shale oil players to find geologic sweet spots and optimize drilling/completion practices by vast correlation-based ML)
GitHub/Physics-Based-Deep-Learning
----- Deep Learning the Physics of Transport Phenomena (This paper generates model data from neural net (NN) without solving PDE, just training on thousands of parameter models and associated expensively-simulated FD model data. Final state only predicted by NN, not intermediate time step data. “[the NN] outperforms the FD model across all possible grid sizes and the computational performance gain improves as the grid size is increased” (10^3 speed up @ 512x512))
List of Machine Learning conferences 2019-20 (tryo labs)
----- another list (aideadlines)
----- another list (wikicfp)
Machine Learning: What it is and why it matters (SAS)
Neural Network
----- Loss landscape visualizer
----- Neural network zoo
----- Visualization of structure, weights and biases
reddit/MachineLearning
TensorFlow
----- GitHub/easy-tensorflow (nothing easy about it)
Tour of convolutional networks
Towards Data Science Blog
Uber AI blog
X lines of Python: Machine Learning (Agile Geoscience Blog; rare example that is well documented, short, and it works)
Why Python for ML?
----- Draw a neural network diagram with matplotlib

----- scikit-learn (flow_chart, glossary)

----- PyTorch
----- ----- Getting started: PyTorch 60 min blitz
----- ----- From TensorFlow to PyTorch, reddit thread

Top Machine Learning Academic Units

US News & World Report: Artificial Intelligence
----- ----- likely US-bias, check out Oxford UK and ETH Germany

----- 1) Carnegie Mellon ML, Course: Intro to ML, KNN demo (very cool)

----- 2)  MIT Artificial Intelligence

----- 3) Stanford Artificial Intelligence

Applications outside geophysics
Big Data and Artificial Intelligence Analytics in Geosciences: Promises and Potential (GSA)

Image registration

Machine learning for data-driven discovery in solid Earth geoscience (Science)

Machine Learning in Geoscience: Riding a Wave of Progress (AGU)

What can machine learning do for geoscience? (AusiMM)

Competitions
AgileGeoscience 2019 Norway
Facebook-Pytorch hackathon