Not sure why it's take so long but Andrew Leigh's excellent "Randomistas" is finally out in the US: https://t.co/QTSbiml0I8 - very good, readable book about randomised trials in economics and the social sciences.
— Tim Harford (@TimHarford) August 6, 2018
This is such a good example of a "wicked hard" data science problem (one that sounds trivial but turns out to be quite complex in practice). https://t.co/pQq41zGxOf
— Hilary Mason (@hmason) August 6, 2018
Prediction: Any AI problem that you can simulate and sample endlessly many training samples for can be solved with today's algorithms such as deep and reinforcement learning.
— Richard (@RichardSocher) August 6, 2018
Fun post: Optimizing floor plans using genetic algorithms. #MachineLearning #DataSciencehttps://t.co/FhLhZ45VDj pic.twitter.com/2PTG6GcO6v
— Randy Olson (@randal_olson) August 6, 2018
A major step in our understanding of OpenAI Five: https://t.co/ki0AiB0BIK
— OpenAI (@OpenAI) August 6, 2018
Neural Beatbox? RNN-based Rhythm Generation + Audio Classification = FUN! https://t.co/b2yisOG0x1 pic.twitter.com/3YYWmMzK68
— Nao Tokui (@naotokui_en) August 6, 2018
This article combines two of my favorite things: an explanation of why randomized controlled trials are the only way to go and a demonstration of how corporate wellness programs totally suck.https://t.co/wGx185J4Q0
— Cathy O'Neil (@mathbabedotorg) August 6, 2018
“The generative predictive model, posits that our brain’s primary function is to reduce surprise by developing an increasingly nuanced model of the world. It constantly predicts the future, noticing incorrect predictions, and updating its model to explain away the difference.” https://t.co/uOjExCZhpN
— hardmaru (@hardmaru) August 4, 2018
Outstanding technical podcast on NLP papers by @nlpmattg and @waleed_ammar of @allen_ai https://t.co/UhBLwY9DfS
— Oren Etzioni (@etzioni) August 3, 2018
Optimum decision making in presence of uncertainty comes from probabilistic thinking. The relevant probs. are of a predictive nature: P(the unknown | the known). Thresholds are not helpful and are completely dependent on the utility/cost/loss function. https://t.co/Z4UNlyWZPx
— Frank Harrell (@f2harrell) August 3, 2018
For anyone using Altair with JupyterLab: last week's JupyterLab 0.33 release comes with an updated Vega-Lite renderer with a newly-styled options menu pic.twitter.com/6svGX7v6jo
— Jake VanderPlas (@jakevdp) August 2, 2018
New blog post: definitely the last time I'm going to go on about the New England Journal of Medicine's statistical guidelines.https://t.co/bwGKTuFnT8
— Simon Gates (@ProfSimonGates) August 2, 2018
