Preprint
Aug 2026
Clustered Randomized Smoothing for Stochastic Prediction Functions
Clustered $\alpha$-smoothing is proposed, a framework that partitions noisy samples using an arbitrary clustering algorithm, applies $\alpha$-smoothing locally within each cluster, and combines the resulting predictions into a mixture distribution, derived from interpreting the smoothing distribution as a mixture of $\alpha$-smoothers.
Eduardo Figueiredo, Frederik Baymler Mathiesen, J. Schumann et al.
· 0 citations