Sparse Function-space Representation of Neural Networks

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.

arXiv commentAccepted to ICML 2023 Workshop on Duality for Modern Machine Learning, Honolulu, Hawaii, USA. 4 pages, 2 figures, 1 table