The Mimo architecture: estimating predictive uncertainty through independent subnetworks

This Github repository contains the code of our project where we reproduce the results of the paper "Training independent subnetworks for robust prediction" by Havasi et al.

Authors: Lorenzo Mazza, Fernando Gastón and Anass Elyasini