AI solves problems by finding patterns in data. AI can only play chess because patterns emerge from the board, pieces, and rules staying the same game after game. Markets are different. So how should investors who want to add value using AI go about it? A THREAD 👇🏼

Dreaming of Thinking Machines; The Dream, The Brain, The Machine. Real though provocation by my friend @QuantsPortal quantsportal.com/dreaming-think…

Impressive! They use #NeuralNetworks to predict the params of a Weibull dist describing P(churn), not churn events: ragulpr.github.io/2016/12/22/WTT…

Total Recall; not a typical "Quant book" but inspiring and filled with lessons nonetheless :-) audible.co.uk/pd?asin=B009GL… h/t Arnold

This perl: print "prime" if(1x$a)!~/^1?$|^(11+?)\1+$/; from github.com/MinhasKamal/Cr… is too cool! q(❂‿❂)p

Here’s the complete #TFDevSummit playlist for those who want to rewatch a talk, or if you missed the livestream. goo.gl/xb561f

If you're interested in Google's approach to business and their commitment to #MachineLearning you should read this: audible.co.uk/pd?asin=B004VF…

Meta-models in scikit-learn are awesome and the docs are better than most University textbooks. Kudos scikit-learn #MachineLearning

Stunning! Visualizations like these which communicate the intuition behind "complex" algorithms are essential for business & user adoption. x.com/astrobiased/st…

Random-Walk Bayesian Deep Networks: Dealing with Non-Stationary Data - twiecki.github.io/blog/2017/03/1…. Very neat Thomas Wiecki! Time to learn PyMC3 :-D

Some neat visualizations which explain the intuition behind statistical concepts students.brown.edu/seeing-theory/ (! mobile friendly) #DataScience

Mind-blowing essays like this - written by leading minds to demystify the cutting edge of science - is why IAS ideas is my favourite site 🖖 x.com/the_IAS/status…

As TensorFlow's #java API matures and smartphones improve, apps will learn from and adapt to their surroundings research.googleblog.com/2017/04/federa… :-D

This is mind blowing! Unsupervised Sentiment Neuron from OpenAI -google.co.za/amp/s/blog.ope… #MachineLearning #NeuralNetworks

Very few could weave the best of Fama, Shiller, and Darwin together as coherently as Prof Lo does: bloomberg.com/news/audio/201…. H/t Barry Ritholtz

LOL! Brilliant, but only 1/3 true. What I also do: cry about (2) how messy historical data is and (3) how long hyperparameter tuning takes! x.com/kaggle/status/…

Great post, Kullback-Leibler Divergence Explained: countbayesie.com/blog/2017/5/9/…. #MachineLearning #Statistics

Can #machinelearning automate the design of neural networks? Learn more about an approach we call "AutoML" goo.gl/gclgdf

A really interesting article about a new math that might answer some of the most interesting questions of our time. x.com/sfiscience/sta…

Learning to Reason with Neural Module Networks bair.berkeley.edu/blog/2017/06/2…