doktormike/evidentialflux.jl

Evidential Deep Learning Layers for Flux

I'm Mike: Father, Scientist, AI developer, Entrepreneur

• 🔭 I’m currently working on artificial general intelligence
• 🌱 I’m currently learning uncertainty estimation using energy based model formalism
• 👯 I’m looking to collaborate on anything bringing our world to a more sustainable and fair place for everyone
• ⚡ Fun fact: I'm not a very funny guy

My latest Videos

• Desupervised - Get Inspired - Commercial Navigator Insurance
• Desupervised - Get Inspired - Commercial Navigator
• Dsupervised - Get Inspired - What is Bayesian AI?
• Desupervised - Get Inspired - AI Levels
• Desupervised Company Intro

My latest Blog posts

• LogP, LogD, pKa and LogS: A Physicists guide to basic chemical properties
• What does the prior actually do in a Bayesian analysis?
• COVID-19 in Denmark - An epidemic in a small country
• A quick introduction to derivatives for machine learning people
• The importance of context

evidentialflux.jl's Issues

Implement corrections to DER

This is from a recent paper on Arxiv. It provides corrections for the loss function as well as the equations for epistemic and aleatoric uncertainty.

Create a comparison between VI and DER

I would like to see a head to head comparison of the uncertainties produced by Variational Inference and Deep Evidential Regression for a moderately complex neural network.

This could be anything related to regression which is non trivial.

Implement the evidence function for the deep evidential regression case

Implement deep evidential classification

TagBot trigger issue

This issue is used to trigger TagBot; feel free to unsubscribe.

If you haven't already, you should update your TagBot.yml to include issue comment triggers.
Please see this post on Discourse for instructions and more details.

If you'd like for me to do this for you, comment TagBot fix on this issue.
I'll open a PR within a few hours, please be patient!

Add additional distributions to Deep evidential regression

Currently we assume that all regression cases will want to use a y~Normal(mu, sigma) model. This is definitely not always true and there may be many other distributions which could be useful for a deep learning setting. For instance.

• Poisson
• Binomial
• NegativeBinomial
• StudentT (This might be a bit flaky since we do end up with a model evidence following a StudentT either way)
• Cauchy