Proximity-Based Evidence Retrieval for Uncertainty-Aware Neural Networks
This provides a more reliable and interpretable method for operational uncertainty-aware decision-making, particularly in domains requiring transparency and auditability, though it is incremental as it builds on existing uncertainty-aware frameworks.
The paper tackles the problem of uncertainty-aware decision-making in neural networks by proposing an evidence-retrieval mechanism that uses proximal exemplars and Dempster-Shafer theory to create instance-adaptive thresholds, resulting in higher or comparable performance with fewer confidently incorrect outcomes on CIFAR-10/100 datasets.
This work proposes an evidence-retrieval mechanism for uncertainty-aware decision-making that replaces a single global cutoff with an evidence-conditioned, instance-adaptive criterion. For each test instance, proximal exemplars are retrieved in an embedding space; their predictive distributions are fused via Dempster-Shafer theory. The resulting fused belief acts as a per-instance thresholding mechanism. Because the supporting evidences are explicit, decisions are transparent and auditable. Experiments on CIFAR-10/100 with BiT and ViT backbones show higher or comparable uncertainty-aware performance with materially fewer confidently incorrect outcomes and a sustainable review load compared with applying threshold on prediction entropy. Notably, only a few evidences are sufficient to realize these gains; increasing the evidence set yields only modest changes. These results indicate that evidence-conditioned tagging provides a more reliable and interpretable alternative to fixed prediction entropy thresholds for operational uncertainty-aware decision-making.