Ax and BoTorch leverage probabilistic models that make efficient use of data and are able to meaningfully quantify the costs and benefits of exploring new regions of problem space. In these cases, probabilistic models can offer significant benefits over standard deep learning methods such as neural networks, which often require large amounts of data to make accurate predictions and don’t provide good estimates of uncertainty.

love the pictorial explanation. everything is better explained with a picture!

"rationality is not knowing facts, but knowing which facts should be considered"

the second example blows my mind. never seen more clear proof that human minds {don't work with, aren't primed for, suck at} probability.

we're all Bayesians now

"This experiment therefore shows that, at least for local models of quantum mechanics, we need to rethink our notion of objectivity. The facts we experience in our macroscopic world appear to remain safe, but a major question arises over how existing interpretations of quantum mechanics can accommodate subjective facts.

Some physicists see these new developments as bolstering interpretations that allow more than one outcome to occur for an observation, for example the existence of parallel universes in which each outcome happens. Others see it as compelling evidence for intrinsically observer-dependent theories such as Quantum Bayesianism, in which an agent's actions and experiences are central concerns of the theory."

accepted papers at:
BayesOpt 2017
NIPS Workshop on Bayesian Optimization
December 9, 2017
Long Beach, USA

We introduced ROBO, a flexible Bayesian optimization framework in python. For standard GP-based
blackbox optimization, its performance is on par with Spearmint while using the permissive BSD
license. Most importantly, to the best of our knowledge, ROBO is the first BO package that includes
Bayesian neural network models and that implements specialized BO methods that go beyond the
blackbox paradigm to allow orders of magnitude speedup.

Snoek 2012 paper
code here: http://www.cs.toronto.edu/˜jasper/software.html
TPA: https://github.com/jaberg/hyperopt/wiki

DeepMind's paper on bayesian optimization

overview of bayesian optimization by the author of Spearmint