Abstract

Deep neural networks (NNs) are known to lack uncertainty estimates and struggle to incorporate new data. We present a method that mitigates these issues by converting NNs from weight space to function space, via a dual parameterization. Importantly, the dual parameterization enables us to formulate a sparse representation that captures information from the entire data set. This offers a compact and principled way of capturing uncertainty and enables us to incorporate new data without retraining whilst retaining predictive performance. We provide proof-of-concept demonstrations with the proposed approach for quantifying uncertainty in supervised learning on UCI benchmark tasks.
Original languageEnglish
Number of pages6
DOIs
Publication statusPublished - 2023
MoE publication typeNot Eligible
EventWorkshop on Duality Principles for Modern Machine Learning - Honolulu, United States
Duration: 29 Jul 202329 Jul 2023
https://icml.cc/virtual/2023/workshop/21496

Workshop

WorkshopWorkshop on Duality Principles for Modern Machine Learning
Country/TerritoryUnited States
CityHonolulu
Period29/07/202329/07/2023
Internet address

Fingerprint

Dive into the research topics of 'Sparse Function-space Representation of Neural Networks'. Together they form a unique fingerprint.

Cite this