Bayesian Deep Learning and a Probabilistic Perspective of Model Construction (Intro + Resources)
Bayesian deep learning is a field at the intersection between deep learning and Bayesian probability theory. It offers principled uncertainty estimates from deep learning architectures. These deep architectures can model complex tasks by leveraging the hierarchical representation power of deep learning, while also being able to infer complex multi-modal posterior distributions. Bayesian deep learning models typically form uncertainty estimates by either placing distributions over model weights, or by learning a direct mapping to probabilistic outputs.
Source: Alex Kendall, Deep Learning Is Not Good Enough,
We Need Bayesian Deep Learning for Safe AI
For the newcomers who want to kickstart Bayesian Deep Learning, you can refer to a great blog post by Joris Baan: A Comprehensive Introduction to Bayesian Deep Learning and the paper Hands-on Bayesian Neural Networks — a Tutorial for Deep Learning Users by Jospin et al.
PAPER
Bayesian Deep Learning and a Probabilistic Perspective of Model Construction, the actual paper can be viewed below or found here.
SLIDES
The slides of the ICML Tutorial can be accessed here.
VIDEOS
The playlist of ICML 2020 Tutorial on Bayesian Deep Learning and a Probabilistic Perspective of Model Construction can be found here:
The tutorial of the previous Bayesian Deep Learning series held at NeurIPS 2019 can be found here:
If you would like more AI/DL resource compilation, do subscribe to my YouTube channel and some claps to keep me motivated!
Support and Donation can be done via Paypal and Patreon (under development), or via Crypto.
BTC ⇢ bc1q2r7eymlf20576alvcmryn28tgrvxqw5r30cmpu
ETH ⇢ 0x58c4bD4244686F3b4e636EfeBD159258A5513744
Doge ⇢ DSGNbzuS1s6x81ZSbSHHV5uGDxJXePeyKy
Cheers!
