Luigi Acerbi

1/ I love Ted Chiang's writing, but about this well-written piece remember that (a) most of the current theories of the brain are that the brain is to a large degree a prediction engine; and (b) compression and prediction are *literally* the same thing.

A shout-out to the excellent lecture series on Foundations of Deep Reinforcement Learning by @pabbeel: highly recommended view for any student of machine learning and also computational neuroscientists. YouTube:

Wow, it's giftmas! This book by Roman Garnett is an amazing resource for everyone working with Gaussian processes, surrogate modeling, active learning, and - you may guess - Bayesian optimization. Looking forward to go through it in detail, and an instant recommendation!