Retrospective Uncertainties for Deep Models using Vine Copulas

Despite the major progress of deep models as learning machines, uncertainty estimation remains a major challenge. Existing solutions rely on modified loss functions or architectural changes. We propose to compensate for the lack of built-in uncertainty estimates by supplementing any network, retrospectively, with a subsequent vine copula model, in an overall compound we call Vine-Copula Neural Network (VCNN). Through synthetic and real-data experiments, Figure 1: VCNN: We propose a plug-in vine-copula module we show that VCNNs could be task (regression/classification) that can complement any neural network with uncertainty and architecture (recurrent, estimates, any time after a model has been trained, without fully connected) agnostic while providing reliable requiring any modifications to it. Additionally, our intervals and better-calibrated uncertainty estimates, capture both - aleatoric and epistemic uncertainty.